Memoir '44 D-Day Landings Memoir '44 D-Day Landings

Forums

Search
Forums » T2R Competitive Play - English » Change of observation mode discussion
Show: Today's Posts 
  
AuthorTopic
erps
Senior Member

User Pages
Posts: 1637
Registered:
July 2005
  Change of observation mode discussion Wed, 20 September 2006 09:08
Hi

I moved this topic from another thread to a new one:

The question is: How to deal with cheating danger and improve observation mode.

Dandy schrieb am Wed, 20 September 2006 04:08

IMHO, the problem with this solution (hide destinations and vagons at player's hand) is that it will actually make oservation quite useless and dull. If you can't see destination and vagons, you can't try to think, analyze and predict the next move, you can't understand the logic for the game and the way of thinking of the player whom you try to observe.

For me it will JUST KILL observation - almost the same as closing the game.

I still insist that saving the game and allow other people observe RECORDED game after the game is finished would be much better experience.

I do not agree with ERPS that observation in real time, but without destination tickets and set of vagons is better than observing recorded game. It will be just very, very dull - imagine, for the first half of the game you probably won't see ANY ACTION AT ALL! Board is static at the beginning. And then when you finally see the action - you won't understand it, because you don't see tickets and vagons. Horrible! That would be like observing soccer - sorry, football - game, but without showing the ball. You will see just random movements of the players, trying to guess, but to no purpose - WHERE IS THIS F...ING BALL??? Laughing

I would partially agree with ElSoy - DoW, your better don't do anything with observation -- if you can't record the game. (Actually, recording the games should be quite simple for you).

To ERPS - recording and watching TTR games would be very interesting even after the game is over!

Same as with chess games - you take the record of the past games and enjoy the great thinking of Morphy, Lasker, Euve, Fisher or Kasparov - no matter how long time ago their beautiful games were played! But you need FULL RECORD to understand the game, so you need to record destinations and vagons in TTR. I would loke to analyze my own mistakes and losses - get the record of my lost game and observe it from the other side would be just FABULOUS!



Okay, little misunderstanding. Of course NOT watching the ticks and colors is NOT better than seeing the whole thing, but i believe it is FAR BETTER than seeing ... NOTHING Smile

Read my proposal careful:
NEW Observation mode: As before BUT even better because you can watch both sides (or all in Multi).
NEW Closed mode: Every game is watchable, but not the ticks and cards and the personal chat in game -> better too, because you can see all games to some limit.
NEW record mode: Store the games for later watching (your proposal)

And we have all we all wanted...

Here the pictures of concept again:

New closed game:
http://www.informatik.uni-frankfurt.de/~erps/zugumzugbilder/screen_new_closed.jpg

New observable game:
http://www.informatik.uni-frankfurt.de/~erps/zugumzugbilder/screen_new_observe.jpg
      
Corflu
Member

Posts: 34
Registered:
July 2006
Re:Change of observation mode discussion Wed, 20 September 2006 13:39
If games are observable, I feel you should be able to see everything for one player (your choice of player) to learn. Whomever starts the game can choose the option for observable or not.

However I feel for something like the Nations Cup, games should NOT be observable. This stops potential cheating. There are many other chances for a "learning environment". Those games are not a learning environment. While others may love to watch and cheer, the game between the two is the thing. What about the players that do not want others to watch their games to learn their style and "scout" them?

per erps talk suggestion: Talking while observing a game sounds nice, but may create a problem due to possible "assistance".
      
Leleline_J
Member

Posts: 88
Registered:
October 2005
Re:Change of observation mode discussion Wed, 20 September 2006 16:49
RB_Corflu schrieb am Wed, 20 September 2006 13:39

If games are observable, I feel you should be able to see everything for one player (your choice of player) to learn. Whomever starts the game can choose the option for observable or not.

However I feel for something like the Nations Cup, games should NOT be observable. This stops potential cheating. There are many other chances for a "learning environment". Those games are not a learning environment. While others may love to watch and cheer, the game between the two is the thing. What about the players that do not want others to watch their games to learn their style and "scout" them?

per erps talk suggestion: Talking while observing a game sounds nice, but may create a problem due to possible "assistance".


I think, that when the games aren't observable anymore, the whole feeling of NC would be gone and it would be boring... It is fun watching the other games and cheer for the teammembers, just playing isn't that fun Twisted Evil Twisted Evil

The players who do not want others to watch their games, shouldn't play in NC.

      
Dandy
Member

Posts: 99
Registered:
June 2005
Re:Change of observation mode discussion Thu, 21 September 2006 05:39
Thanks to erps for starting this new thread - otherwise the whole idea of recording the TTR games with a possibility to replay them would be buried under heated exchange of accusations.

After reading carefully the last erps reply I see that there is full consensus in our opinions: "reduced" observation mode in real time is better than nothing, "full" observation mode in real time is more interesting for observer, but not secure from cheaters and potentially creates a lot of tension for players if it is not simply mutual agreement by players, but FORCED on the playes by the rules of NC tournament.

At the same time, I agree with Leleline_J (and disagree with Corflu) - indeed, if "the games aren't observable anymore, the whole feeling of NC would be gone and it would be boring... It is fun watching the other games" - I will second that opinion.

Still, we should try to avoid even the sightest possibility for cheaters to compromise the whole idea of NC Tournaments. How we can reconcile observation and protection from cheating?

The only solution I can see - recorded games and delayed observation.

That will preserve the public nature of tournaments and at the same time will guarantee fair play. Yes this compromise won't allow cheering in real time mode - but all interesed parties will be able to observe the game immediately after it, comment on it, share thei views, suggest better moves - there could be, for example, public observation mode, where players could exchange their opinions, and observers could ask questions and actively participate in discussion and commenting.

Hey, folks, this could be even much MORE interesting than simple silent observation in real time! Public observation with active participation for observers and comments from the players!
      
SKMorefield
Senior Member

User Pages
Posts: 621
Registered:
January 2005
Re:Change of observation mode discussion Thu, 21 September 2006 13:35
My guess is recorded games won't happen unless the person playing or an observer records the game. I'm not sure I like the idea anyway because I don't even like to watch a recorded football game where I know the outcome. All the tension is lost.

Part of the fun of NC is watching games. However, there is the fear of cheats. And, cheating isn't as easy to detect as one might think. Just the knowledge of tics is enough to sway the outcome even if you don't purposefully block (thus making it obvious you know the tics). For instance, if you have LA-Miami and you see that your opponent has 2 small tics, you will be more careful in your route than you would if he/she had Van-Mont. For a cheater, cheating can be entirely undiscovered, and yet happen every game!

So, the solution? As erps already said, make full observation an option along with partial observation. This way, people who want to see an entire game with tickets and all still can do this, yet players can still CHOOSE to partially allow their games to be observed, showing the board only. This would be the same as if someone walked up on a table with the cards down and saw the tracks. There is still something to be gained.

The only difference now is top players play closed... so how is that accomplishing anything?? Something is better than nothing!



SKM
      
Corflu
Member

Posts: 34
Registered:
July 2006
Re:Change of observation mode discussion Thu, 21 September 2006 15:05
Yet if this option is not introduced by DoW, I feel the NC games should be closed. Contrary to what FS-Leleline_J thinks, the competition would be just as fierce and enjoyable by the teams. Though less fun for the fans. But consider the risk and acusations that will be avoided.
      
erps
Senior Member

User Pages
Posts: 1637
Registered:
July 2005
Re:Change of observation mode discussion Thu, 21 September 2006 15:38
Hi
RB_Corflu schrieb am Thu, 21 September 2006 15:05

Yet if this option is not introduced by DoW, I feel the NC games should be closed. Contrary to what FS-Leleline_J thinks, the competition would be just as fierce and enjoyable by the teams. Though less fun for the fans. But consider the risk and acusations that will be avoided.


You think the accusations will stop? No chance if you let the people decide if they play open or closed (or semi closed it's the same). The only solution for that is stating in the rules: All games must be played close! But there are enough players like bassie and me who mostly don't care if they are watched esp. against each other. And if two players play against each other who know the other player and trust him 100% they will play open. In any other case they play closed and this means (i hope you can see the logic) that they not trust each other, so it is a hidden accusation... At least the other player can think it is a hidden accusation even if other reasons exists...

I don't want ALL closed games in NC as a rule and i think the last years problems with open/close were the reason for bassie doing ONLY open games. He hoped and me too, that the advantage of no discussion is higher than the disadvantage of some cheaters. Now we see, that we still have discussion because players find a way to avoid being watched and we still have the disadvantage of possible cheaters Sad

I think seeing the tracks in a closed game (semi closed how ever) helps a lot to get a little of the live feeling but we need the possibility to watch an entire game too. I hope DOW will change the closed mode soon, together with communicating the ticks to your friends it is exciting enough.

There is still another (more complex) solution:
A player who wants to watch your game could be able to "ask" you. Let's say a message is showing with "player Anu_ would like to watch" and you are able to say yes or no. Of course a nuisance in a NC game with 30 potential observers! Advantage: It is possible to watch one of both sides if you can choose the player of your desire. You can limit the "ask" time frame to such like 30 seconds or one minute. All questions are stored and presented at once in a pop up screen where you click some buttons for "all", "none", "buddys" or single names. It's possible but really not as easy as the other solutions.

bye, erps

      
Corflu
Member

Posts: 34
Registered:
July 2006
Re:Change of observation mode discussion Thu, 21 September 2006 18:25
erps wrote on Thu, 21 September 2006 09:38


You think the accusations will stop?



Yes, if the games are closed.

erps wrote on Thu, 21 September 2006 09:38


But there are enough players like bassie and me who mostly don't care if they are watched esp. against each other. And if two players play against each other who know the other player and trust him 100% they will play open.



The problem is that everyone is not erps and bassie Razz And then anyone who does not play open would be indirectly acusing the other as possibly cheating. Let's eliminate any potential insult.


erps wrote on Thu, 21 September 2006 09:38


There is still another (more complex) solution:
A player who wants to watch your game could be able to "ask" you. Let's say a message is showing with "player Anu_ would like to watch" and you are able to say yes or no.



A great suggestion! Though then you have the issue of accepting some and not others. Plus it can be tedious and bothersome when playing to be asked the question if someone joins mid-game. Not to mention what if you allow some and not others. I like the though though!



Personally I LOVE watching the games and competition and learn a lot from it. And I would really like to watch any replays afterwards. Yet I think the potential for unfair play must totally be eliminated. I wish we lived in a society where we could all be trusted to be honest and fair. The sad truth is the few that don't act this way ruin it for all. If we were all honest in the world, there would be little need for a police force.

[Updated on: Thu, 21 September 2006 18:29]

      
GANG_dea1
Senior Member
T2R Nation Cup 2007 Winner

Posts: 1856
Registered:
September 2005
Re:Change of observation mode discussion Thu, 21 September 2006 23:31
RB_Corflu schrieb am Thu, 21 September 2006 18:25

erps wrote on Thu, 21 September 2006 09:38


You think the accusations will stop?



Yes, if the games are closed.



nope

Games closed and not recorded?
J'accuse ...
- player A played on against bot for more than one move (that discussion sounds familiar, doesn't it?)
- player B deliberately slowed down the game (not so new either)
and nobody can prove it, because there is no witness

OK, so all games shall be recorded - that implies DOW does it (noone else can, as the games are closed) and stores them all on some biiiiiig server. Rolling Eyes We can hardly oblige all tournament participants to record all their own games, can we? Rolling Eyes
For the sake of the argument let's assume the recording is accomplished somehow:

J'accuse ...
- player C has played with player D's account (yes, we've had that too)
- player E has played his match while players F,G,H,I were sitting next to him, discussing the best move (ok, I just invented that one - haven't seen that in the forum yet)

It won't get any better whatever we do.
People, who really want to cheat, will find a way.
People, who want to hunt witches, will find/create witches.
I enjoy playing TTR because I trust the other players (and nobody gave me a reason to doubt his honesty as of yet).
If there comes a day when I stop to believe in my opponent's honesty all I can do is stop playing.

RB_Corflu schrieb am Thu, 21 September 2006 18:25


erps wrote on Thu, 21 September 2006 09:38


There is still another (more complex) solution:
A player who wants to watch your game could be able to "ask" you. Let's say a message is showing with "player Anu_ would like to watch" and you are able to say yes or no.



A great suggestion! Though then you have the issue of accepting some and not others. Plus it can be tedious and bothersome when playing to be asked the question if someone joins mid-game. Not to mention what if you allow some and not others. I like the though though!



Maybe this solution is easier:
Choice between
- closed
- open to everyone
- only people on your buddy list can watch
combined with erps' idea to be able to select, which player you want to watch, of course.

If someone normally not on your buddy list wants to watch he can ask beforehand. No disturbance during play.

Just a suggestion - I'll continue playing open anyway.

Regards
dea
      
Leleline_J
Member

Posts: 88
Registered:
October 2005
Re:Change of observation mode discussion Fri, 22 September 2006 16:59
dea1 schrieb am Thu, 21 September 2006 23:31


Games closed and not recorded?
J'accuse ...
- player A played on against bot for more than one move (that discussion sounds familiar, doesn't it?)
- player B deliberately slowed down the game (not so new either)
and nobody can prove it, because there is no witness

OK, so all games shall be recorded - that implies DOW does it (noone else can, as the games are closed) and stores them all on some biiiiiig server. Rolling Eyes We can hardly oblige all tournament participants to record all their own games, can we? Rolling Eyes
For the sake of the argument let's assume the recording is accomplished somehow:

J'accuse ...
- player C has played with player D's account (yes, we've had that too)
- player E has played his match while players F,G,H,I were sitting next to him, discussing the best move (ok, I just invented that one - haven't seen that in the forum yet)

It won't get any better whatever we do.
People, who really want to cheat, will find a way.
People, who want to hunt witches, will find/create witches.
I enjoy playing TTR because I trust the other players (and nobody gave me a reason to doubt his honesty as of yet).
If there comes a day when I stop to believe in my opponent's honesty all I can do is stop playing.




I totally agree with dea. I think, that there isn't a really good solution for this cheater-problem. Everybody record his own game and all games are closed ?? not possible.

Just the people of the buddy-list can watch the game ? for me a bad idea.

I know so many nice people who I trust, but not everybody is on my buddylist. and I think that's quite unfair for the other people, because they haven't the possibility to watch. and one of the reason that we play open games is as well, that the lower ranked people can learn of the games of good players. and then they couldn't, because they aren't in the buddylists of good players if they don't know them. and it's right, that it is annoying when you try to concentrate during your game and you always get ask, if somebody could watch it. Sometimes there are around 20 people who watch a NC-game.

All games half-closed or open, is may the best idea.

But why not just trusting the other people??? It's right, that there are people who cheat and lie. But gets the world better if everybody mistrust to the other ????? NO. and here it's just a game!!! It's not as important as the REAL life outside of online-games !!
if somebody needs to cheat (mostly because he want to have RESPECT of the other people and that he can feel GOOD, because he won a game against a top-player) he WILL cheat, if the games are open or closed. Like dea said, every cheater finds his way. But mostly, good players will notice, if somebody really cheats... and then they don't trust this player anymore and they won't have RESPECT of him. so maybe the cheater will win the NC or just some games(but I don't think so, a player who cheats is mostly not a top because otherway he wouldn't have necessary to cheat, he would win anyway. and a real good player can win a cheater), but the other players are distrustful and nobody would play with him again. And I think, the longer we speak about cheater or not cheater, the more cheater will appear...

That's just my opinion.

Since I am playing I had just once the guess that somebody cheats. But that's over and I trust all players. I will always play open.

Greetings
Leleline_J

[Updated on: Fri, 22 September 2006 17:30]

      
Goan
Member

User Pages
Posts: 99
Registered:
July 2005
Re:Change of observation mode discussion Fri, 22 September 2006 18:05
I totally agree with Joli, Bas, Erps, El Soya etc.

I realy do not understand this, in my view, paranoide fear of cheating. What's the real problem ?

Ok, once in a while, especially when you play regularly against unknown players, you faces a cheater who has i.e. multiple accounts) and perhaps you loses the game by that. Ok 10 minutes of your life has been wasted. Compared to the time and energy you spent to try to avoid cheaters, put hunderds of postings to ventilate your opinion about it, lobbying and begging for closed games, make (false) accusations to others, annoyes other people. It is not the end of the world, is it ?

Ah, it is about your rating ? Hmm. If you trust the rating system (you do because otherwise you would not pay attention to it), that same system will place you on the right spot again in the long run. The "Archilles and the Turtle"-adepts are right that it does matter a bit (0.01 points or so) but hey..

I also simply do not believe that a good player cheats. He is simply much to proud to try to win by cheating. Winning like that does not do much for his ego. Furthermore, I cannot understand that a good player thinks that a second good player thinks that the first one cheats.

The NC is maybe a different story. In here a complete team is involved. But on the other hand, cheating on those games can be done by other means. One can for example let him advice his moves by his (better) team-members. Not that I believe this happens btw. What I want point by this is that cheating is not abandoned by closing the games (partly or completely).

There could be a valid reason for closing ones games. Sometimes you do not want to be watched by your next opponent in the NC (or you do not want to reveal your excellent technique...). But to me this is a mere luxury which I personally can live without.

So get rid of this Close/Open Option. As far as I know, the open NC matches are really an advertisement for this game. You will never see more players online than during this tournament. That's the merit of the games being open. I think no one could argue against that.

And Yes. Of course I also do not like cheaters and cheating. Of course you should try to abandon that. But not to the cost of everything.

Goan

Btw. We have a saying in Holland: "Zoals de waard is, zo vertrouwt hij zijn gasten". It means something like "Like the Innkeeper is, so he trusts his guests". I am not saying this applies to the T2R-players who are panickally mistrust cheaters. But it is something one can think about.
      
Dandy
Member

Posts: 99
Registered:
June 2005
Re:Change of observation mode discussion Sat, 23 September 2006 00:27
dea1 wrote on Thu, 21 September 2006 14:31


OK, so all games shall be recorded - that implies DOW does it (noone else can, as the games are closed) and stores them all on some biiiiiig server. Rolling Eyes We can hardly oblige all tournament participants to record all their own games, can we? Rolling Eyes
For the sake of the argument let's assume the recording is accomplished somehow




I want to clarify my idea about RECORDING TTR GAMES - by DoW:

I do NOT suggest to make a VISUAL recording - as we currently do for our private recordings of TTR games. These recordings are indeed huge video files, and you can't do much with such record.

What I actually suggested (and, by the way, ERPS got it right from the very beginning!) - is to make a recording of MOVES send & received by DoW server.

When we play TTR, we are not getting any pictures from TTR Server - we are getting short instrictions what happened in the game, and the same goes in the other direction - when you make a move, your computer do not send any pictures, it sends internal representation of your move, which is a VERY SHORT MESSAGE.

These messages, that age going between clients and server, constitute a full record of the game, and such record should be very short and efficient - you don't need huge servers and a lot of disk space to keep such short records. Of course, visual recordings are absolutely not suitable for the purpose.

Our clients (A.K.A "Computer TTR Game") should be able to interpret such recordings and replay the game. Also, it would be nice if out Clients could make our private recordings of our own games and store them locally on our computers. Again, these recordings will be very short files, made of bit messages, not visual recording of the screen action!

DoW folks, can you consider adding this recording functionality to your server and to the next version of Computer TTR Game?

Thanks,
Dandy (Software Engineer)
      
GANG_dea1
Senior Member
T2R Nation Cup 2007 Winner

Posts: 1856
Registered:
September 2005
Re:Change of observation mode discussion Sat, 23 September 2006 02:27
Dandy schrieb am Sat, 23 September 2006 00:27


I do NOT suggest to make a VISUAL recording


Don't worry, Dandy - nobody does.
But even in a world of bits which can either be 0 or 1
a lot of 0 is almost as much as a little bit of 1 Wink

Sure - it would be great to have that - but still - work and storage on the DOW side (unless you only follow the idea of storing locally - that one's great but won't help for NC).

Best regards
dea (yet another software engineer)
      
Dandy
Member

Posts: 99
Registered:
June 2005
Re:Change of observation mode discussion Sun, 24 September 2006 20:09
dea1 wrote on Fri, 22 September 2006 17:27

Dandy schrieb am Sat, 23 September 2006 00:27


I do NOT suggest to make a VISUAL recording


Don't worry, Dandy - nobody does.
But even in a world of bits which can either be 0 or 1
a lot of 0 is almost as much as a little bit of 1 Wink

Sure - it would be great to have that - but still - work and storage on the DOW side (unless you only follow the idea of storing locally - that one's great but won't help for NC).

Best regards
dea (yet another software engineer)



Glad to meet a workfellow! Then there is no need to explain to each other anything about 0 and 1, bits and bytes - excellent!

As a software engineer, you shold see that such recordings will be tiny files, not taking disk space. Actually, they should be even shorter than the messages in this forum, at least some of them! Razz

My estimation for standard TTR game:

Number of moves to take vagons - 40-50 moves for each player,
Number of moves to claim tracks - 12-20 moves,
OK, add a few moves to pick more destination tickets for "ticket fiends" Razz -
in total it should not be more than 75 moves per player anyway!

This gives us 150 moves per 2-player game - say, 4 bytes to encode the move, that should be enough - 600 bytes at maximum, plus a few bytes more to record initial state of the game - open vagons, 4 vagons at player's hands and initial destination tickets...

My estimation: standart 2-3 player game should take 0.5 - 1 KB, 4-5 player game - 1 - 2 KB.

This should not be an issue to keep short history of all games played - say, a week history would be... how many games? ... another estimation - if we see ~30-50 simultaneous games, a game should take 10-15 min - approx 200 games per hour played, maybe 4000 games per day, less than 30,000 games per week - 60,000KB = 60MB - that is tiny volume! I can keep a full history of all TTR games ever played on Earth on my home computer! (you can make this estimation yourself and figure out the size of hard disk on my home computer - as an exercise Razz )

No need for a "biiiiiig server" from you previous reply Laughing and any significant "storage on the DOW side " - that is, if they do it the right way.

Dandy
      
Brice
-= Crew =-
Commandant

User Pages
Posts: 1917
Registered:
October 2002
Re:Change of observation mode discussion Mon, 25 September 2006 11:35
Dandy wrote on Sun, 24 September 2006 20:09


No need for a "biiiiiig server" from you previous reply Laughing and any significant "storage on the DOW side " - that is, if they do it the right way.

Dandy


Sorry to enter this discussion lately. I won't discuss the background of the topic, just the above sentence.

You seem to confuse desktop hardware class and enterprise hardware class.
Storing 60MB per day (as per your computation, which is yours not mine (since I didn't calculted it)) on a standard desktop hard drive might be cheap.

Unfortunately you forgot that this information will be written almost always in parallel. You talked about 30-50 games in parallel (which imho is underestimated), this means 30~50 threads writing to the hard drive at the same time and moreover there will be some user replaying those games stored.

I'm confident that desktop hardware won't be able to handle this (and I'm not talking about possibility of hard drive crash, etc...). And if it handles it, I'm pretty sure it won't be scalable (ie what happens if you have a sudden surge in traffic and you must handle 200 parallel games??).

Now transpose it in 'enterprise' terms and it means a powerfull RAID10 array with 15k RPM drives and we're not at all in the same price range (and that's the minimum, because the movements will certainly be hosted in a database to be searchable, or browsable and that's adding a new layer).

That's one of the reason we don't offer this feature (that doesn't mean we won't offer it in the future).
      
Dandy
Member

Posts: 99
Registered:
June 2005
Re:Change of observation mode discussion Wed, 27 September 2006 03:52
Brice wrote on Mon, 25 September 2006 02:35

Dandy wrote on Sun, 24 September 2006 20:09


No need for a "biiiiiig server" from you previous reply Laughing and any significant "storage on the DOW side " - that is, if they do it the right way.

Dandy


Sorry to enter this discussion lately. I won't discuss the background of the topic, just the above sentence.

You seem to confuse desktop hardware class and enterprise hardware class.
Storing 60MB per day (as per your computation, which is yours not mine (since I didn't calculted it)) on a standard desktop hard drive might be cheap.

Unfortunately you forgot that this information will be written almost always in parallel. You talked about 30-50 games in parallel (which imho is underestimated), this means 30~50 threads writing to the hard drive at the same time and moreover there will be some user replaying those games stored.

I'm confident that desktop hardware won't be able to handle this (and I'm not talking about possibility of hard drive crash, etc...). And if it handles it, I'm pretty sure it won't be scalable (ie what happens if you have a sudden surge in traffic and you must handle 200 parallel games??).

Now transpose it in 'enterprise' terms and it means a powerfull RAID10 array with 15k RPM drives and we're not at all in the same price range (and that's the minimum, because the movements will certainly be hosted in a database to be searchable, or browsable and that's adding a new layer).

That's one of the reason we don't offer this feature (that doesn't mean we won't offer it in the future).



Brice,
First of all - big thanks for paying attention and responding to this thread!

I quite agree that my desktop won't be able to replace DoW server - not in my wildest dreams! I'm talking about only one additional feature, and the additional workload and volume on hard drives needed on DoW server - that is all.

Brice, you over-estimate additional traffic and resources needed for this feature, in my opinion.

First of all, you don't need to write all this information on hard drive in parallel - since you don't need to share this info immediately, and the size of data is negligible, each thread can store it in a buffer, and another thread can receive completed buffers, ready to be written on disk. This should be done only after the game is over. That should be quite scalable solution even for a peak load - unless TTR community will grow exponentially. Smile

Additional disk space for storing games is not big at all - not more than 20GB of data generated per year, according to my estimation. And you probably do not need to keep all games indefinitely - you can delete obsolete stuff older than a month (or a week, if you like) from general store, and create a few special, permanent libraries for Tournaments, NC Cups and Brilliant Immortal Games - nomination and selection of the games could be done by Tournament Directors or by Community Librarian. I'm sure that you can find volunteers for this post! Razz

Thanks again for your reply - and especially for that last sentence and the words in parenthesis - "that doesn't mean we won't offer it in the future". It gives me some hope!

Dandy
(Software Engineer, Microsoft)
      
erps
Senior Member

User Pages
Posts: 1637
Registered:
July 2005
Re:Change of observation mode discussion Wed, 27 September 2006 07:51
Hi,

Dandy schrieb am Wed, 27 September 2006 03:52


Additional disk space for storing games is not big at all - not more than 20GB of data generated per year, according to my estimation. And you probably do not need to keep all games indefinitely - you can delete obsolete stuff older than a month (or a week, if you like) from general store, and create a few special, permanent libraries for Tournaments, NC Cups and Brilliant Immortal Games - nomination and selection of the games could be done by Tournament Directors or by Community Librarian. I'm sure that you can find volunteers for this post! Razz


Store them one week but make the downloadable for all users. In this way you can use THEIR Desktop machines for storage Wink Then add another feature to the java client: Opening stored games on your harddrive for replay. Or add it at least to the pc game Smile Another reason for buying!!

Dandy schrieb am Wed, 27 September 2006 03:52


Dandy
(Software Engineer, Microsoft)


I'm really sorry for you Very Happy

bye, erps
      
Zeno
Senior Member
Cadet

User Pages
Posts: 581
Registered:
February 2006
Re:Change of observation mode discussion Wed, 27 September 2006 19:01
I love, and hate, the idea of having a record of the games. On the one hand, it sounds like a wonderful thing but ...

i) Privacy concern: I think you only have a right to have a record of games you have played in.

ii) Competetive concern: Compile the data and you can have a pretty good idea of what a person will do based on the cards they play, and the cards you have. Of course the good players do this already, but it relies on intuition and experience. Give me more information, and better information, and I can turn this into a science instead of an art. Moreover, the competitive balance swings to the number-crunching, algorithm-producing, computer-programming types (who seem to be pursuing this).

My suggestion: at the end of the game have the option to have a receipt of the game showing names, tickets drawn and taken, cards face up, and what action was performed each turn. This will be sent as a PM to you, hence stored in space that is already alloted to you. (I do not know if the present Java program keeps track of the history or only the present position, so this may not be feasible.) You only get receipts for games you are in. The code could be a simple CSV file, though it should be readable by a patch to the PC/Mac game. It is up to you to save the games locally, and empty your PM space from time to time. These receipts could be traded like baseball cards, and tourneys could mandate that receipts be stored somewhere (DoW's, shamogi's and/or erps' site), but for the private/paranoid types, we might even want to specify that closed games do not produce receipts.

Just my reaction to a very good, but potentially dangerous, idea
      
Dandy
Member

Posts: 99
Registered:
June 2005
Re:Change of observation mode discussion Fri, 29 September 2006 05:08
Zeno wrote on Wed, 27 September 2006 10:01

I love, and hate, the idea of having a record of the games. On the one hand, it sounds like a wonderful thing but ...

i) Privacy concern: I think you only have a right to have a record of games you have played in.

ii) Competetive concern: Compile the data and you can have a pretty good idea of what a person will do based on the cards they play, and the cards you have. Of course the good players do this already, but it relies on intuition and experience. Give me more information, and better information, and I can turn this into a science instead of an art. Moreover, the competitive balance swings to the number-crunching, algorithm-producing, computer-programming types (who seem to be pursuing this).

My suggestion: at the end of the game have the option to have a receipt of the game showing names, tickets drawn and taken, cards face up, and what action was performed each turn. This will be sent as a PM to you, hence stored in space that is already alloted to you. (I do not know if the present Java program keeps track of the history or only the present position, so this may not be feasible.) You only get receipts for games you are in. The code could be a simple CSV file, though it should be readable by a patch to the PC/Mac game. It is up to you to save the games locally, and empty your PM space from time to time. These receipts could be traded like baseball cards, and tourneys could mandate that receipts be stored somewhere (DoW's, shamogi's and/or erps' site), but for the private/paranoid types, we might even want to specify that closed games do not produce receipts.

Just my reaction to a very good, but potentially dangerous, idea


Privacy concern: There is simple solution, similar to observation - players who played in the game should have the right to allow or prevent publishing the recorded game, but participants should be always able to get the recording of their own games anyway, published or not published. There could be special policy for tournaments though - say, tournament games should be always published.

Competetive concern: I do not believe anybody would really like to create his own "improved bots" instead of just playng the game themselves. And since we are not bots, our moves are not that predictable. Yes, possibility to look at the games and analyse it (just using our own brain) will be good for a lot of players - that is why strong players do a lot of observation. But I believe it is OK - as a reslut, there will be more strong players, more interesting games, better understanding of the game.

My favourite analogy - with chess: strong chess players do not object to the fact that their games are written and published in chess databases and could be analysed. Why should TTR players be so scared?

Zeno suggestion: I'm not that interested in statistics - and I believe most players won't need that much statistical data about the games. I know this is different for Zeno - but this is due to his personal addiction to statistics. Razz Playing the game, observing the game, recording, and later replaying and analyzing the game (but only with my own brain, not with any computer!!!) - yes, that should be interesting for many players, not just me - at least, I think so...

Dandy
      
    
Previous Topic:strange behavior
Next Topic:Multiplayer Ranking Top 12
Goto Forum: