Round 3 is up!
http://www.slitherine.com/tourn_results/1
Good luck everyone!
Does anyone have any feedback so far on how the system is working?
Beta tournament Round 3
Moderators: Slitherine Core, FoG PC Moderator, NewRoSoft
-
deadtorius
- Field Marshal - Me 410A
- Posts: 5290
- Joined: Mon Oct 20, 2008 12:41 am
So far things have gone smoothly for me, but I can see problems with the time limits and real life. My work has picked up unexpectedly so I personally might find it difficult to make the game turns in time and I do recall someone posting they were going to be away for work during round 4. Not sure how to work around that one. The Pyrrhus scenario was a good choice, it was fast and deadly. My opponent and I got it done in 1 day 
-
ianiow
- Lieutenant Colonel - Panther D
- Posts: 1235
- Joined: Fri Apr 27, 2007 11:24 am
- Location: Isle of Wight, UK
I am enjoying the tournament so far. No technical problems to report. I do have some requests/observations though:
1) When the two battles appear in my inbox, I would like to know which battle is which (what side I am playing).
2) Awarding 50 pts for a bye is waaay too much. Not even Pantherboy can manage to score 50 point per round. Maybe a bye should score an average of the other winning scores for that round?
3) The Draws table is confusing to read. I would prefer seeing the Score1 column renamed Battle1 and have both players score in it for that battle.
Keep up the good work!
1) When the two battles appear in my inbox, I would like to know which battle is which (what side I am playing).
2) Awarding 50 pts for a bye is waaay too much. Not even Pantherboy can manage to score 50 point per round. Maybe a bye should score an average of the other winning scores for that round?
3) The Draws table is confusing to read. I would prefer seeing the Score1 column renamed Battle1 and have both players score in it for that battle.
Keep up the good work!
-
CheerfullyInsane
- Staff Sergeant - Kavallerie
- Posts: 302
- Joined: Sat Aug 14, 2010 12:11 pm
- Location: Birkerød, Denmark
Aside from the E-mail boo-boos things have run smoothly.
(Btw, am I the only one who got a second letter stating that the third round had started?)
I'd second ianiow's opinion that the battles should state which side you're playing.
It's not that big a deal since you'll figure it out when opening the game, but it would still be nice.
Then there's the whole time-limit thing.
Admittedly, there are problems due to people having a life outside FoG, hard as that is to imagine.
But there's no real solution to it in my opinion. Yes, you run the risk of losing because of other obligations, but the whole point of the time-limit is having tourneys end at a specific time.
Finally, there's the infamous 50pts gained on a time-forfeit.
It does seem like a lot of points. On the other hand, if it was anything less than the maximum possible points, you run the risk of people manipulating their games, losing on time by choice.
Example: Say there's a rule that says you gain only as many points as the average of all other game-winning scores in that round.
Now, you've hammered your opponent into a bloody pulp, and he's set to lose the game by 40 pts if it ended right now.
But no-one else has won the scenario by more than 30 pts so far in the round.
Now you're opponent can simply run out the clock, and lose by max. 30 pts, even though he's far worse off.
Obviously, in a perfect world nobody would think like this, but.....
So, all in all I think the system is great idea. Presumably it'll support other formats than Swiss at some point, but I can wait for that.
(Never really liked the Swiss system myself.)
Regardless, it'll save some of our current tourney organizers a *lot* of work once it gets up and running.
Lars
(Btw, am I the only one who got a second letter stating that the third round had started?)
I'd second ianiow's opinion that the battles should state which side you're playing.
It's not that big a deal since you'll figure it out when opening the game, but it would still be nice.
Then there's the whole time-limit thing.
Admittedly, there are problems due to people having a life outside FoG, hard as that is to imagine.
But there's no real solution to it in my opinion. Yes, you run the risk of losing because of other obligations, but the whole point of the time-limit is having tourneys end at a specific time.
Finally, there's the infamous 50pts gained on a time-forfeit.
It does seem like a lot of points. On the other hand, if it was anything less than the maximum possible points, you run the risk of people manipulating their games, losing on time by choice.
Example: Say there's a rule that says you gain only as many points as the average of all other game-winning scores in that round.
Now, you've hammered your opponent into a bloody pulp, and he's set to lose the game by 40 pts if it ended right now.
But no-one else has won the scenario by more than 30 pts so far in the round.
Now you're opponent can simply run out the clock, and lose by max. 30 pts, even though he's far worse off.
Obviously, in a perfect world nobody would think like this, but.....
So, all in all I think the system is great idea. Presumably it'll support other formats than Swiss at some point, but I can wait for that.
(Never really liked the Swiss system myself.)
Regardless, it'll save some of our current tourney organizers a *lot* of work once it gets up and running.
Lars
I've got two words for ya: Math is hard.
Been very smooth for me too.
I like the system.
I am not too worried about the title differentiation but agree that the
I also think the potential 50 points for a bye too much on the other hand Cheerfully Insane's point is well made.
The halfway house which is still simple might be:-
The player awarded the bye gets average winning points for each of the mirrors the bye applies to - the player the bye is awarded against can only ever get zero.
Time limit is troublesome for me this w/e due to work commitments but I can see no other solution to running a tight completion schedule other than to have the timelimit.
I like the system.
I am not too worried about the title differentiation but agree that the
I also think the potential 50 points for a bye too much on the other hand Cheerfully Insane's point is well made.
The halfway house which is still simple might be:-
The player awarded the bye gets average winning points for each of the mirrors the bye applies to - the player the bye is awarded against can only ever get zero.
Time limit is troublesome for me this w/e due to work commitments but I can see no other solution to running a tight completion schedule other than to have the timelimit.
I would like to see a separate tab in the Multiplayer window dedicated to tournament games.
Also, when rewarding a forfeit due to time limit, give the remaining player the average score of all his victories until that point, or the average victory points for all the other players if it is the first round.
I would also like see the option for tournaments with different rules (like in a knock-out tournament, which I believe would be much more popular than the current one's system or a groups+knock-out one).
Also, when rewarding a forfeit due to time limit, give the remaining player the average score of all his victories until that point, or the average victory points for all the other players if it is the first round.
I would also like see the option for tournaments with different rules (like in a knock-out tournament, which I believe would be much more popular than the current one's system or a groups+knock-out one).
-
IainMcNeil
- Site Admin
- Posts: 13558
- Joined: Fri Apr 01, 2005 10:19 am
Thanks for the feedback guys, it is very valuable.
* Score 1 means player 1 score, not the score from Battle 1. It is the sum of their scores from both games. Can you think of a better header? Just score for both? We want to keep the headers as narrow as possible to avoid issues on low res monitors.
* On time limits we're already looking to change the way it works to make it less intensive. You'll still have a timie limit but if you run out of time, you will only default the game if it does not complete before the next round kicks off. This means more games will get completed and being away for a weekend is less of an issue as you can pick up and continue the game during the week. This should be sorted for the next tournament. This should mean far less games time out.
* On BYE scores I agree and had already discussed internally. There are 2 issues here - where to insert the player in the draw and what score to allocate. average score was the idea I was going with but its an issue when you are in round 1 or 2. We really need to take the average score at the end of the competion and fit it back. This is a little messy. If we do this we need a way to give them a score for the BYE round for the purposes of ranking to get an opponent. This is a bit tricky to resolve so wont be fixed in the immediate future but thoughts on how best to do it are welcome. I think a full 50 for where the plaeyr defaults on time is probably ok - its the games where no game is played I want to address.
* Drop out - we are adding a system to support drop out. If a player wants to they can resign. If a player does nothing for an entire round this will be dropped out automatically. This way any inactive players will be cut from teh draw to avoid the messed up games in the future round. This will not deal with someone missing a single round and wanting to continue but to be honest I cant think of a fair way to allow this. If you miss a round I think you have to be taken out of the comp.
* For the game name in the Inbox it might be tricky because of limited space. A new tab for tournament games might help here. That wont happen in the near future though.
* What are the e-mail boo boos Cheerful?
Although we want to add to and impriove the system over time is there anything so far you have seen that would prevent us from launching to a wider audience or any quick fixes that woudl really help enhance other than what is mentioned above?
* Score 1 means player 1 score, not the score from Battle 1. It is the sum of their scores from both games. Can you think of a better header? Just score for both? We want to keep the headers as narrow as possible to avoid issues on low res monitors.
* On time limits we're already looking to change the way it works to make it less intensive. You'll still have a timie limit but if you run out of time, you will only default the game if it does not complete before the next round kicks off. This means more games will get completed and being away for a weekend is less of an issue as you can pick up and continue the game during the week. This should be sorted for the next tournament. This should mean far less games time out.
* On BYE scores I agree and had already discussed internally. There are 2 issues here - where to insert the player in the draw and what score to allocate. average score was the idea I was going with but its an issue when you are in round 1 or 2. We really need to take the average score at the end of the competion and fit it back. This is a little messy. If we do this we need a way to give them a score for the BYE round for the purposes of ranking to get an opponent. This is a bit tricky to resolve so wont be fixed in the immediate future but thoughts on how best to do it are welcome. I think a full 50 for where the plaeyr defaults on time is probably ok - its the games where no game is played I want to address.
* Drop out - we are adding a system to support drop out. If a player wants to they can resign. If a player does nothing for an entire round this will be dropped out automatically. This way any inactive players will be cut from teh draw to avoid the messed up games in the future round. This will not deal with someone missing a single round and wanting to continue but to be honest I cant think of a fair way to allow this. If you miss a round I think you have to be taken out of the comp.
* For the game name in the Inbox it might be tricky because of limited space. A new tab for tournament games might help here. That wont happen in the near future though.
* What are the e-mail boo boos Cheerful?
Although we want to add to and impriove the system over time is there anything so far you have seen that would prevent us from launching to a wider audience or any quick fixes that woudl really help enhance other than what is mentioned above?
-
ianiow
- Lieutenant Colonel - Panther D
- Posts: 1235
- Joined: Fri Apr 27, 2007 11:24 am
- Location: Isle of Wight, UK
iainmcneil wrote:
* Score 1 means player 1 score, not the score from Battle 1. It is the sum of their scores from both games. Can you think of a better header? Just score for both? We want to keep the headers as narrow as possible to avoid issues on low res monitors.
I understand why you have done this. My suggestion is to put battle1 in the first column with player A's score always on the left of the '/' and player B's score always on the right, and batttle2 score in the other column:-
_______battle1__battle2
ianiow__11/14___10/15__deeter
(Here deeter has won both games)
It just seems more logical (the way the english football scores are reported!)
I'm glad to have won both fictional battles, so will add my tupence worht. I don't think the current scoring is good. Actually winning a game isn't rewarded, only attrition is. I also think, as each round starts on a Friday, time should not run out until Thursday night.
Having said that, so far I like the beta for two reasons: 1. it uses scenarios which means a chance to play with realistic armies in realistic situations unlike current DAG matches with their gamy orders of battle and totally weird maps. 2. the auto-challenges are very convenient. The current league games are a bit cumbersome having to track down 15 other player's PMs per division.
I do hope this takes off and gets perfected because DAG games often yield impossible-to-win army/terrain matchups and so aren't a true test of generalship, but rather ogf gamesmanship. It is interesting to note that formerly invicible players are now losing often because, in the current league season, they aren't allowed to use their invicible armies.
Deeter
Having said that, so far I like the beta for two reasons: 1. it uses scenarios which means a chance to play with realistic armies in realistic situations unlike current DAG matches with their gamy orders of battle and totally weird maps. 2. the auto-challenges are very convenient. The current league games are a bit cumbersome having to track down 15 other player's PMs per division.
I do hope this takes off and gets perfected because DAG games often yield impossible-to-win army/terrain matchups and so aren't a true test of generalship, but rather ogf gamesmanship. It is interesting to note that formerly invicible players are now losing often because, in the current league season, they aren't allowed to use their invicible armies.
Deeter
-
IainMcNeil
- Site Admin
- Posts: 13558
- Joined: Fri Apr 01, 2005 10:19 am
Good point - I forgot to mention scoring. We are looking for a mechanic to reard the win, as currently you get nothing for taking the enemy army down other than the final break point.
We want a scoring system that encourages aggressive play though so we dont want to penalise the loser too heavily.
We're welcome to suggestions on the best scoring system.
Maybe something like now except that we allocate 20 points instead of 10 between the 2 players. If you break the enemy army you then get a 5 point bonus. The only isue here is that it does nto give you a fixed number of points for a round so its harder to allocate a BYE. We coudl say teh BYE gets the 20 ) score without the 10 point bonus. Its a pretty good score.
An alternative is to only score the % of the enemy army you have killed ignoring your loses. This rewards very aggressive play.
Every scoring system will have its strengths and weaknesses but the current one definitely does not reward breaking the enemy army enough.
We want a scoring system that encourages aggressive play though so we dont want to penalise the loser too heavily.
We're welcome to suggestions on the best scoring system.
Maybe something like now except that we allocate 20 points instead of 10 between the 2 players. If you break the enemy army you then get a 5 point bonus. The only isue here is that it does nto give you a fixed number of points for a round so its harder to allocate a BYE. We coudl say teh BYE gets the 20 ) score without the 10 point bonus. Its a pretty good score.
An alternative is to only score the % of the enemy army you have killed ignoring your loses. This rewards very aggressive play.
Every scoring system will have its strengths and weaknesses but the current one definitely does not reward breaking the enemy army enough.
Ian,
Here is a suggested system that keeps things close but can reward aggression and/or sound strategy.
Simple bog standard baseline scores
Win/Bye = 3 points
Draw = 1 point
Loss = 0 points
However reward aggression thus :-
Bonus points +1 for each increment of 5AP* difference between final AP scores for winner.
Penalty points -1 for each increment of 5AP* difference between final AP scores for looser.
Simple - it rewards aggression, punishes loosing badly (and not fighting to 'bitter end') but the rewards* are not so extensive that having a bye (3 points only) rules you out of contention. (*The suggested 5AP could be larger e.g. 10AP to further control bonuses and keep 'bye' players in closer contention).
You could also reward speed of victory with a bonus point to any winner with fastest (fewest turns) victory in any given round.
Here is a suggested system that keeps things close but can reward aggression and/or sound strategy.
Simple bog standard baseline scores
Win/Bye = 3 points
Draw = 1 point
Loss = 0 points
However reward aggression thus :-
Bonus points +1 for each increment of 5AP* difference between final AP scores for winner.
Penalty points -1 for each increment of 5AP* difference between final AP scores for looser.
Simple - it rewards aggression, punishes loosing badly (and not fighting to 'bitter end') but the rewards* are not so extensive that having a bye (3 points only) rules you out of contention. (*The suggested 5AP could be larger e.g. 10AP to further control bonuses and keep 'bye' players in closer contention).
You could also reward speed of victory with a bonus point to any winner with fastest (fewest turns) victory in any given round.
keithmartinsmith will win the tournament. who wants to bet with me on this?!!
why? simple.. because Archie was awarded 50 points for his opponent leaving the tournament, which actually is more gained in 1 round that he gained in the other 2 previous round added up. and as the score board looks at this moment, he'll play keithmartinsmith, while I'll play pantherboy. now, considering the previous results and what we know about the players, keith will nullify Archie, while me and panther will most probably have a close match.
ergo, keith has the most chances to win the tournament. does the matchup for the first 4 places seem right? of course not. it will be played and solved like above? most probably, yes.
An abandoned match should be awarded with the average score of all the the games played that round. it won't be perfect, but I think it should be the simplest and fairest way to do it. how should the math look like? assuming 6x2 matches per round, from which 4 pairs ended up as 1-1 and 2 pairs ended up as 2-0 means most probably an abandoned pair should have ended as a 1-1 match (ie one victory + one defeat). then compute the average victory points from all matches played that round, and average defeat points, and award the player with those. All this math for the first round only. If it happens on the second round, you can enhance the algorithm by comparing remaining player's previous round(s) performance vs previous round(s) average performance, computer current rounds average performance and modify it accordingly with the difference computed for the previous round(s).
a very simple and yet fair way to do it.
why? simple.. because Archie was awarded 50 points for his opponent leaving the tournament, which actually is more gained in 1 round that he gained in the other 2 previous round added up. and as the score board looks at this moment, he'll play keithmartinsmith, while I'll play pantherboy. now, considering the previous results and what we know about the players, keith will nullify Archie, while me and panther will most probably have a close match.
ergo, keith has the most chances to win the tournament. does the matchup for the first 4 places seem right? of course not. it will be played and solved like above? most probably, yes.
An abandoned match should be awarded with the average score of all the the games played that round. it won't be perfect, but I think it should be the simplest and fairest way to do it. how should the math look like? assuming 6x2 matches per round, from which 4 pairs ended up as 1-1 and 2 pairs ended up as 2-0 means most probably an abandoned pair should have ended as a 1-1 match (ie one victory + one defeat). then compute the average victory points from all matches played that round, and average defeat points, and award the player with those. All this math for the first round only. If it happens on the second round, you can enhance the algorithm by comparing remaining player's previous round(s) performance vs previous round(s) average performance, computer current rounds average performance and modify it accordingly with the difference computed for the previous round(s).
a very simple and yet fair way to do it.
-
CheerfullyInsane
- Staff Sergeant - Kavallerie
- Posts: 302
- Joined: Sat Aug 14, 2010 12:11 pm
- Location: Birkerød, Denmark
Sorry for the late response, I completely forgot about this thread.iainmcneil wrote:* What are the e-mail boo boos Cheerful?
Although we want to add to and impriove the system over time is there anything so far you have seen that would prevent us from launching to a wider audience or any quick fixes that woudl really help enhance other than what is mentioned above?
The e-mail thingie referred to the mail sent out when you were doing in-house testing, so we got a message stating we had already started on Wednesday (IIRC).
Also I got a duplicate letter for Round 3 a couple of hours after the first.
No biggies, and I'd suspect it was a mail-server error rather than a flaw in the tourney itself.
On a side-note, I mentioned I didn't like the Swiss system. Which isn't entirely correct.
I just think that the Swiss system is far better suited for ranked play, so that players are handicapped when entering the tourney.
In this one I think I went from 23rd to 5th in one round, since first I played pantherboy, next what I suspect is another newbie.
Now, these things will obviously sort themselves out as the rounds progress, and the league has already started to shake itself out, but for the Swiss system to work one either has to have ranked entry, or more than 4 rounds.
For open play I'd prefer either knock-out, or round-robin.
But again, no biggie. This is a beta, and the system itself works beautifully. Other formats can wait until we get scoring settled.
Lars
PS: ian, I might have missed this in an earlier post, but is this system to be for Slitherine alone, or will it be so that players can construct their own automated tourneys?
I've got two words for ya: Math is hard.
-
davouthojo
- Sergeant First Class - Elite Panzer IIIL
- Posts: 423
- Joined: Wed Mar 24, 2010 1:49 pm
- Location: Hong Kong
I like the system, it is a good addition to FOG.
The swiss system rankings...I echo the comments above - have a pure knockout alternative, and any issues with Swiss can be resolved through ranked entry and/or more turns.
The timer....I don't like it, but can't think of anything better. Certainly I would give the players as much time as possible, right up to the start of the next round. I would set the next round to begin as soon as all matches from the previous round are complete, so no dead time between rounds. Several good formulas are quoted above. I think zero for the player running out of time is OK. The problem is 25 for the winner/bye.
Cothyso's average works for most cases. The only add is if the game has advanced a long way before the timer runs out, you could use the BP score in the match the moment the clock runs out? If you did this, plus a bonus for the player not running out of time, it would reflect the maximum information on the players ability. It would be unfair to give a player on the verge of a stunning victory an average result because his opponent ran out of time. You could give them the higher of this calculation, or the average.
Scoring...you need to reward the win more. Currently, once both sides reach 75% BPs, there is little to play for
Strongly recommend Slitherine opens the system to players so we can design and recruit for our own tournaments....andersm73's job would be much easier with this software....
The swiss system rankings...I echo the comments above - have a pure knockout alternative, and any issues with Swiss can be resolved through ranked entry and/or more turns.
The timer....I don't like it, but can't think of anything better. Certainly I would give the players as much time as possible, right up to the start of the next round. I would set the next round to begin as soon as all matches from the previous round are complete, so no dead time between rounds. Several good formulas are quoted above. I think zero for the player running out of time is OK. The problem is 25 for the winner/bye.
Cothyso's average works for most cases. The only add is if the game has advanced a long way before the timer runs out, you could use the BP score in the match the moment the clock runs out? If you did this, plus a bonus for the player not running out of time, it would reflect the maximum information on the players ability. It would be unfair to give a player on the verge of a stunning victory an average result because his opponent ran out of time. You could give them the higher of this calculation, or the average.
Scoring...you need to reward the win more. Currently, once both sides reach 75% BPs, there is little to play for
Strongly recommend Slitherine opens the system to players so we can design and recruit for our own tournaments....andersm73's job would be much easier with this software....