|
| TWC predictions | 2008-04-22 15:11:19 |
Is it possible to guess the TWC winner?
The list of teams sorted by average rating (2 of April) looks as follows:
Team..AvgRating...Games Played Last ... |
| Ando Meritee | 2008-04-22 18:27:47 |
It is interesting to compare ratings. But there is a problem in your comparison method because some teams consist of 4 members, some 5. More proper comparison would be:
1. To compare 4 highest ratings of each team only. In that case, Estonia's average rating is 2442.75. (Each team has a right to not use the 5th player at all, for example Estonian team in TWC 2004).
2. Or to make an average from 5 players, while for the teams of 4 players, the 5th player rating is 0 (it is fair because if a blank substitutes a player, its sure loss), then divide the sum of ratings by 5. |
| |
| Dmitry Epifanov | 2008-04-22 22:56:31 |
Team...AvRt......Games...RtOn5......RtBest4
Chi......2428,5.....73.........1942,8.....2428,5
Rus1..2411,4.....429........2411,4....2442,5
Est.....2400........417........2400.......2442,75
Jap.....2340,75..225........1872,6....2340,75
Swe...2215,2...185.........2215,2....2235,5
Rus2..2194,2....276........2194,2....2212
Fin.....2058,25..68...........1646,6...2058,25
Cze....1987,25..220........1589,8...1987,25
Ukr....1980,25...65..........1584,2...1980,25
Really hard to format such tables =(
I suppose that both best-4 and five-0-if-no-fifth average ratings are poor. The first is bad because when you have a possibility to give rest to a tired player, to make a substitution placing uncomfortable opponent etc. your team will perform better. The second is bad because the necessity of this 5th player is not 100%. How to calculate it before-hand obviously, I don't know.
The only thing I can offer is calculating average rating based on the quantity of games played. If 1st play 100% his games, 2nd 100%, 3rd 80%, 4th 70% and 5th 50% (for 3rd and 4th), it is easy to calculate. But this can be done only afterwards when there will be no need in predictions =)
So... Maybe you can try to predict the result, Ando? |
| |
| Ando Meritee | 2008-04-22 23:03:17 |
| Easy to predict - Estonia will be the 1st of course. :) Russia and China will fight for the 2nd and 3rd place. Sweden and another Russian team will fight for the places 4-5. Japan will be the 6th. The rest of the places belong to Finland, Czech and Ukraine. |
| |
| Ando Meritee | 2008-04-22 23:20:45 |
About the comparing best 4 ratings of each team - I do not understand your logic. Your first method was the average of 5, whereas the 5th player is the lowest ranked and makes the average rating smaller compared to the teams of only 4 players. If you compare the situation in real tournament, where a team can choose the substitute player to play, then it means, his strength in specific condition is higher than the player who is being replaced. So, we can fairly then say that he deserves at least same rating as the player who is being replaced. Otherwise, why need to replace at all if it does not improve situation.
We can assume that all players have equal chance to become tired during tournament, or become sick, etc. A team of 4 players cannot use the possibility to substitute player even if the team member for some reason (for example sickness, phobia, etc) becomes weaker than a substitute player.
So if a team with a substitute player can "improve" the team in certain condition, while the team of only 4 players cannot improve the team, then it is illogical to use the average rating of 5 players of one team in comparison with teams of 4 players only. The 5th player deserves at least same rating as the player he is going to replace. For the sake of fairness, we can say he deserves the 4th player rating at least. So, in fact, the average rating of the team of 5 players should be "the average of 4 best + a little more". But your calculation makes the average of the whole team to be less than the average of best 4 players. |
| |
| Dmitry Epifanov | 2008-04-22 23:57:41 |
Te usual practice is to set to substitute not the weakest player but a middle-strength player. So really this formula is 'some (not best) 4 players + sth extra'. This team's performance may be even worse then '4 best players'.
And if the substitutor is the weakest player, then it is quite optimistic to say that he will substitute the 1st or the 2nd board showing the same performance =)
So.
Well, I agree that 5-players avg rating must be something like '4 players + something extra". That's right, but how to calculate that? It seems that 2 models that you offered are quite poor (read above about that). My model (just avg rating) is also not fair. Maybe it is possible to create better one?..
And anyway. I've created this topic not for holywar on how to calculate team rating (it must also affect the 'team spirit' and much more), but for forecasts.
My forecast - Rus1 Chi Rus2 Est =)))) May the dream come true =) |
| |
| Ants Soosõrv | 2008-04-25 21:42:00 |
| Thanks, Dima! :) But why are so pessimistic about Russia 2 place? :) |
| |
| Dmitry Epifanov | 2008-04-25 22:44:26 |
| Because it seems that Rus2 will play without me =((( |
| |
| Bjorn Lind | 2008-04-26 09:09:10 |
Well, i am quite sure that Sweden will win this tournament.. ;)
Then Russia 1, China and Estonia will fight for place 2-4. Japan and Russia 2 will be 5-6!
What are the odds of my forecast? I guess i will be a millionare if i bet on it and sweden wins this tournament? :) |
| |
| Lijing Sun | 2008-04-26 10:21:35 |
I guess the TWC Champion will be one of teams China, Russia-1 and Estonia, the follow will be Russia-2, Japan, Sweden,
BTW, the abbreviation of China is CHN, but not CHI.
CHI is Chile abbreviation. |
| |
| Ales Rybka | 2008-04-27 20:05:02 |
Hi,
we (czech team) are aghast with your forecast and placing us among the last teams. If you by any chance know there is very very (I have to repeat it again) VERY difficult to get 1st kyu comparing with your 5th dan (of course I know that you were counting rif rating).
Our tip is czech team on the second place, after finland which gets the 1st place.
Ales, Pavel, Petr and Monika
(just kidding :-D) |
| |
| Ando Meritee | 2008-04-27 20:34:05 |
I think there is a certain possibility that Czech team would get even the 1st place. It is when all the renju stones are replaced with the caps of beer bottles. And before a player gets the right to use a cap of a beer bottle as a stone for next move, he has to consume the content of the bottle first. The chances are that the game ends quite easily, because opponent will just be too drunk to finish the game, and he will lose by time.
It is also important to notice that the labels of bottles are replaced with the ones from juice or milk bottles, and a player has to keep a confident pose of a sober condition, because according to rules, "obviously drunk players will not be allowed.....".
I think, in given conditions, Czech and Finland should win this tournament easily. :) |
| |
|