Page 1 of 2

Season 3 Top Player and All-Stars announced...

PostPosted: Thu Mar 10, 2011 3:50 pm
by jpcloet
in the next newsletter......... ;)

In attempt to drive readership up and the clan content, the Season 3 All-Stars and the top player will be announced in the next issue of the newsletter due out any week now.

Clan League: MVPs announced by laughingcavalier (TSM)
What you've all been waiting for: Head Clan Director jpcloet has run the regular season results through his high-powered number cruncher and given us the Most Valuable Players of Clan League Season 3.

First up come the divisional all-stars, the best players by game type in their division. No surprise to see some well-known names from division winning clans KORT, TOFU, Empire and IA, and strong performances from LoW, AoC and (no bias, guys) TSM. Special congrats to the players from clans who show up in these lists less often, notably Theldin of The Last Warriors and Gambler2120 of Angels of Death.

Division MVPS:
Image

Who were the all-star teams across the divisions? These players pretty much all won at least 6 games in their position and scored an 80% win rate. Two players each for TOFU and IA, representatives of Empire and KORT. Specially good performances from relative newcomers - two FOED players danryan and Master Fenrir made it in, and Devil's Brigade are represented by Shaneback.

Image

But there can be only one player who takes the general achievement medal for best performance in the league.

jpcloet wrote:The top player award started with all players above 70% and then I do a further data dive. I've been working on a Clan Rating (somewhat similar to QB ratings) and the criteria for the league was 5 components although the expectations for each category did vary based on the results in each group. I found myself consistently having to raise the bar as there were some really strong performances.

The 5 Categories were: Quads Win %, Triples Win %, Doubles Win %, Unique Maps %, and Kills Ratio %. Along with a minimum of 10 games.

Most of the top players had perfect scores in 3 or 4 of the categories. The Quads scores were very close as well. The biggest difference was kill score. There were some players who won a lot of games but were not active participants in some games. The best player for kill ratio in the top group was actually 4red and the 5 players mentioned below combined were an astounding 83-12 for 87.3% win rate. Another interesting note was that the top player did not qualify for a positional spot since he was 4-0, 4-0, and 4-0.


Image

Big congratulations to all the winners, and especially to steve066, CL3 MVP.

These are the players we all want to beat if we can only figure out how to do it. And to help us figure that out, we asked a few of the contenders to give us their teamgame tips. This is what they said:

dwilhelmi (Empire) wrote:It is important to be on the same page with the rest of your team, and that means constant communication. Even when everybody knows everybody else, and you all have a comfortable style, communication is still important. Even the most experienced players can miss something, but if you have three or four people looking at every move, you are a lot less likely to goof up. Along with that is keeping up on your chats, even when it isn't your turn. Asking for input doesn't do a lot of good if your teammates are not checking games when it isn't their turn.


danryan (FOED) wrote:My general tip would be that communication trumps skill for the larger team games. Although a plan may fall apart due to dice or opponents' play, if the team is communicating, formulating a new plan can bring the battle back with a good result. Knowledge of specialist maps is very helpful as well - there are a number of maps where standard gameplay will get you a loss in team games, including Feudal & Epic, Lunar War, Middle Ages, etc. My last tip is pick good teammates, inevitably they will raise the quality of your game as you don't want to let them down.


josko.ri (KORT) wrote:Clearing regions around a bonus is often more important than taking the bonus.


steve066 (KORT) wrote:My only skill is in picking genius teammates :)

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Thu Mar 10, 2011 4:14 pm
by Gold Knight
jpcloet wrote:in the next newsletter......... ;)

In attempt to drive readership up and the clan content, the Season 3 All-Stars and the top player will be announced in the next issue of the newsletter due out any week now.


I want to know who they are, but just cant bring myself to read anything written by LC... so guess Ill miss out. :lol:

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Thu Mar 10, 2011 6:47 pm
by laughingcavalier
Gold Knight wrote:
jpcloet wrote:in the next newsletter......... ;)

In attempt to drive readership up and the clan content, the Season 3 All-Stars and the top player will be announced in the next issue of the newsletter due out any week now.


I want to know who they are, but just cant bring myself to read anything written by LC... so guess Ill miss out. :lol:


Ha ha ha.

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Thu Mar 10, 2011 7:28 pm
by jpcloet
Follow these easy steps to get a copy of the Dispatch directly to your inbox.

viewtopic.php?f=429&t=117941&start=0

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 6:42 pm
by laughingcavalier
Congratulations to all the players who made the divisional & league best teams, and the one very special Most Valuable Player.
Read all about it:

CC Dispatch [Issue 44: 22-3-11] Curious George(rx7di)

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 6:46 pm
by angola
How were the Division MVPs decided?

Congratulations to everyone.

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 6:48 pm
by jpcloet
angola wrote:How were the Division MVPs decided?

Congratulations to everyone.



The Divional All-Stars are based on position. Overall MVP was described in the pm, so it should be in the article.

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 7:01 pm
by the.killing.44
Classic. 6 TOFU and 3 LoW. Go 2008-2009.

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 7:15 pm
by jpcloet
the.killing.44 wrote:Classic. 6 TOFU and 3 LoW. Go 2008-2009.


How many of the TOFU were in LOW before the split, I didn't think it was all of them. I don't remember Sam812 being in LOW in 2008 either, maybe more time has passed than I thought.

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 9:09 pm
by Rodion
I went to check the 12 divisional triples MVPs because I wanted to see how close I was from getting into the all-star triples.

Got some questions for you, JP. :)

1 - Master Fenrir (nominated) had a 7-1 triples record. So did Gambler21210 and myself. What made MF's 7-1 record more deserving of a spot than Gambler's or mine?

I'm guessing you probably broke the tie by seeing who had better doubles/quadruples records, which in my opinion shouldn't really be applied towards determing the best triples performers, but I can understand that, as you probably lacked better tiebreaking criteria.

Question 1 is sort of moot (I'd still like an official answer, though) because, when I checked all 12 players, I concluded (according to the way I see things) that neither MF nor Gambler nor me should earn that 3rd triples spot. Check JP007. The guy had an 8-1 record (88.88%, +7) and got snubbed by someone with a 7-1 record (87.5%, +6).

2 - Why was JP007 snubbed of his deserved (at least in my eyes) triples all-star spot?

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 9:11 pm
by Rodion
By the way, I have absolutely nothing against Master Fenrir, I'm just calling things the way I see them! ;)

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 9:18 pm
by jpcloet
Check their kill scores......hint hint (active vs passive), also most players had a perfect Map score as well. I really like the scoring system for the league based on the games played per member. My current challenge is how to apply it to a single war.

Here are the top trips in order of score, top 2 had the same score

Clan-Div Player Name
DB-B Shaneback
IA-C Forza AZ
FOED-C Master Fenrir
IA-C Jabajabba
TLW-D Theldin
FOED-C L M S
IA-C Redfelt
KORT-D Rodion
DB-B JP007
AOD-B Gambler21210
DB-B MarshalNey
TSM-D Laughingcavalier
LOW-A Clanlord Carl
NEM-C Pmchugh
LOW-A Samuelc812
AOC-C Slickstm
TOFU-A Velvecarrots
DYN-C KraphtOne
FOED-C Nippersean

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 9:24 pm
by jpcloet
Rodion wrote:I'm guessing you probably broke the tie by seeing who had better doubles/quadruples records, which in my opinion shouldn't really be applied towards determing the best triples performers, but I can understand that, as you probably lacked better tiebreaking criteria.


No other game type was considered so you guess wrong, the 5 categories were listed in the article. It was all Trips Win %, Unique Map Score and Kill Score. There was one instance where I had to go to a tie-breaker in another position and did a combo effect, which moved the score enough of a fraction to make a call.


The 5 Categories were: Quads Win %, Triples Win %, Doubles Win %, Unique Maps %, and Kills Ratio %. Along with a minimum of 10 games.

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 9:29 pm
by Rodion
So, the 8-1 guy (JP007) got bumped from his "tied for 1st position" all the way down to 9th just because the team (which included him) decided that dropping armies on someone else would be smarter for those particular games?

In my opinion, the only way of measuring active vs. passive is checking game chat to see who is really leading the game and who is merely following orders. Of course you can't do that. But making a new criterium ("most kills" or "most kills per game", whichever you used) to determine that is... just... :sick:

Am I a dissenting opinion on this subject?

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 9:32 pm
by Army of GOD
Don't worry about it guys, next season I'll sweep the board.

EDIT: Forgot wink ;)

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 9:58 pm
by jpcloet
Rodion wrote:So, the 8-1 guy (JP007) got bumped from his "tied for 1st position" all the way down to 9th just because the team (which included him) decided that dropping armies on someone else would be smarter for those particular games?

In my opinion, the only way of measuring active vs. passive is checking game chat to see who is really leading the game and who is merely following orders. Of course you can't do that. But making a new criterium ("most kills" or "most kills per game", whichever you used) to determine that is... just... :sick:

Am I a dissenting opinion on this subject?


You may not be. Differentiating between 8-1 and 7-1 can be very difficult. Is 19-2 better than 12-0? Now imagine having 10-15 players in that range as well. The kill score could be part of the drop, some players would be in a better position to be the lead in a game even if they are following orders. I'd love to hear other options on how to differentiate performance that is measurable.

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Tue Mar 22, 2011 10:15 pm
by Rodion
I'd say 12-0 beats 19-2, but I understand arguments can be made for both sides.

Now, 8-1 beats 7-1 100% of the time: it wins both percentage wise (88.88% > 87.5%) and "surplus wise" (+7 > +6). Think of it this way: if you gave the 7-1 guys a chance to play one more game, they could (best case scenario) win it and go to 8-1 (drawing with the original 8-1). They could also lose it (worst case scenario) and drop to 7-2 (losing to the original 8-1). The difference between 8-1 and 7-1 is the 8-1 has 100% chance of winning "game 9" (he done it already), while the 7-1 has less than 100% chance of winning "game 9".

It's pretty hard to differentiate performance. I'm not sure I have good ideas on how to break ties. I'm sure, however, that you don't need to apply tie breakers when there are no visible ties that need to be broken (see 8-1 and 7-1 case).

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Wed Mar 23, 2011 2:04 am
by Denise
I agree with you, Rodion. IMO, kills is a poor indicator of performance and shouldn't be considered. I'm not sure what could be used for the tiebreaker, though. Maybe most amusing chat?

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Wed Mar 23, 2011 4:04 am
by Dako
Kill ratio is totally lame. Most valuable players is the frontier? I have not heard about a clan that loads only one guy because he is better at attacking. It always differs from game to game.

So... kill ratio as an MVP criteria is lame, I think.

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Wed Mar 23, 2011 4:15 am
by jpcloet
Rodion wrote:Now, 8-1 beats 7-1 100% of the time: it wins both percentage wise (88.88% > 87.5%) and "surplus wise" (+7 > +6).


Let me first say, there is a significant amount of data analysis that went in to the criteria. I hope more players raise questions and ideas because this is not easy. I would love to have something that is 100% right all the time, but I'd like to think that my experience tells me that won't happen.

Part 1

I'll counter with an example reminiscent of Season 1.

A player is 10-0, only plays doubles and only plays on 2 maps, and 9 of the games are home games and with the same partner.

B player is 10-1, plays both doubles and triples and plays on 11 unique maps and has more unique partners.

Does A still trump B due to % and "surplus"

(I did not add in unique partners as based on % wise, it would penalize smaller clans and is harder to track)

Part 2

I'll counter with a Season 2 example

Player A is 12-0 overall (game wise is 12-0 on quads)
Player B is 22-2 overall (game wise is 10-0 on quads, 7-1 on trips and 5-1 on dubs)

Does A still trump B due to % and "surplus"

Bruceswar and I have had several discussions on what the win % expectation should be based on game type. I've done some work to prove it out with data, but I probably need to do some more.

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Wed Mar 23, 2011 4:18 am
by jpcloet
Dako wrote:Kill ratio is totally lame. Most valuable players is the frontier? I have not heard about a clan that loads only one guy because he is better at attacking. It always differs from game to game.

So... kill ratio as an MVP criteria is lame, I think.


It was really meant only to differentiate to some extent passive vs active. I will not pick on a certain clan, however, there are players who were say 8-1 and did almost nothing in the game than give their men to another player. With a lot of great records, you need criteria outside of just win % IMO to differentiate.

Thank you for your opinion, however, I'm looking for suggestions as to how to improve. What other criteria would you use?

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Wed Mar 23, 2011 4:43 am
by Dako
Yes, I know you cannot say who was better in 8-1 record between 3 players.

Firstly, dropping your armies on another player can be the only way for the team to win the game. It is a strategy and it doesn't devalue the player.

I'd use another criteria - for example, you can try to implement the score of the opposing team to influence the results. Or check if it was home or away games. As a tie-breaker you can count semi's records.

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Wed Mar 23, 2011 4:48 am
by jpcloet
Dako wrote:I'd use another criteria - for example, you can try to implement the score of the opposing team to influence the results. Or check if it was home or away games. As a tie-breaker you can count semi's records.


How would you compile the scores easily, considering that we are using primarily the tournaments script to get our data?


Here is a clan war example of the top 9 players, which one would you pick for MVP?

Record Win
3-0 100%
7-1 88%
6-1 86%
5-1 83%
6-2 75%
6-2 75%
6-2 75%
5-2 71%
7-3 70%

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Wed Mar 23, 2011 4:54 am
by Leehar
7-1?

Re: Season 3 Top Player and All-Stars announced...

PostPosted: Wed Mar 23, 2011 4:55 am
by jpcloet
Leehar wrote:7-1?
Why would you say 7-1? Was 3 games not enough even though they all got medals?