EPTCareer-to-Date Winning % - Page 2
| Blogs > Mumei |
|
danl9rm
United States3111 Posts
| ||
RaGe
Belgium9944 Posts
On February 01 2011 19:22 mizU wrote: Flash started off as a beast. Jesus. Before Flash had even played a televised game, KTF players (specifically reach afaik) were already mentioning his beastly status in practice games during interviews. It seems like they, despite him being one of the better players, waited to use him till he was fully mentally ready. There's no predicting what a first televised loss could do to the confidence of a 15 year old (or was he 14 at the time?) rising talent. I think the KT managers did a good job letting Flash mature a bit before sending him out. | ||
|
Mumei
United States254 Posts
Click the names for the graphs; using the image tags doesn't seem to work very well. I think Jaedong deserves some sort of award for consistency. Flash Jaedong Bisu Stork Fantasy EffOrt Jangbi iloveoov YellOw NaDa Boxer Reach sAviOr And I did two more for newer players who have been doing well in recent months, Bogus and Hydra, since they were the players that got me interested in starting this whole thing. Bogus + Show Spoiler + Through Month Wins Losses Games Percentage Hydra + Show Spoiler + Through Month Wins Losses Games Percentage Really love seeing how Hydra just exploded in the last two months after about two and a half years of mediocrity. I wonder what changed for him. | ||
|
Kyuukyuu
Canada6263 Posts
It's kind of amusing looking at all of JD's April slumps  | ||
|
Mumei
United States254 Posts
On February 02 2011 03:31 Kyuukyuu wrote: Spoilering each spreadsheet would be nice :x It's kind of amusing looking at all of JD's April slumps Good idea, and done. I've been working on career winning percentage rankings for overall (>100 wins), vT, vZ, and vP (>33 wins), and only the top 20 for both, but since there's obviously been so much history I'm sure I've missed a lot. It also tends to penalize players who have played for years past their prime (like Nada, Reach, Boxer, Yellow, etc.), and I'm probably missing quite a few of them. If I am missing players, I would like to know. Still, it is interesting seeing how T and P tend to dominate vT, T tends to dominate vZ, and Z tends to dominate vP. And it is interesting noting the outliers; Jaedong and Yarnc being the only Zergs in the top 20 in vT; Bisu being the only Protoss in the top 20 against Zerg; Flash being breaking up the Zerg dominance in vP to take first there, and Jaedong doing the same thing in vZ with regards to Terran dominance there. Overall + Show Spoiler + Player Wins Losses Games Percentage vT + Show Spoiler + Player Wins Losses Games Percentage vZ + Show Spoiler + Player Wins Losses Games Percentage vP + Show Spoiler + Player Wins Losses Games Percentage | ||
|
hmsrenown
Canada1263 Posts
| ||
|
Harem
United States11390 Posts
On February 01 2011 21:21 Mumei wrote: The Reach discrepancy is due to counting 2v2 wins along with 1v1. -_- | ||
|
Mumei
United States254 Posts
On February 02 2011 08:31 Harem wrote:The Reach discrepancy is due to counting 2v2 wins along with 1v1. -_- Oh, thanks for the explanation. | ||
|
L0thar
987 Posts
| ||
|
eshlow
United States5210 Posts
On February 01 2011 19:57 patrick321 wrote: Forget flash; nada was a 70% king for the majority of his first 2 years. Boxer gets a lot of credit for being amazing but his stats are surprisingly low. Oov was a beast as well, no surprise. Thanks for the very interesting data! You have ot remember back in 2000 there were only leagues. Thus he was only playing the best of the best for some time in leagues. It's expected that you would have a lower percentage of wins in this case. Once proleague came around they started playing a lot more games against crappier opponents (although they got less prep time so it may even out) | ||
|
RezChi
Canada2368 Posts
| ||
|
Mumei
United States254 Posts
On February 02 2011 23:39 eshlow wrote: You have ot remember back in 2000 there were only leagues. Thus he was only playing the best of the best for some time in leagues. It's expected that you would have a lower percentage of wins in this case. Once proleague came around they started playing a lot more games against crappier opponents (although they got less prep time so it may even out) The odd thing is, according to Liquipedia, Proleague didn't start until March 2003 - and it seems like Boxer, who suddenly went from 63% lifetime to a 50% record between March 2003 - 2004, wasn't one of those players who benefited from Proleague. Possibly owing to that relative lack of prep time that you mentioned? I've also heard that he had trouble keeping up with the macro revolution; perhaps that mattered more when he didn't have prep time. (I'm starting to sound like I'm talking about Batman, aren't I?) I'd be interested in hearing more from someone who was actually around back then. On February 03 2011 03:10 MarineHeroSamKhan wrote: Man from the spreadsheet, you can tell that Stork had Flash's talent, but he wasted it on some shit games (WoW) and other stuff. JangBi just makes me sad = ( He was supposedly the Perfect protoss, a combo or Bisu and Stork... but now = ( To be honest, I'm not really sure I see what you're talking about - how early was Stork's laziness supposed to have started? He seemed consistently below Flash at basically any point I look at - the only time it seemed close was in the first ten - twenty-ish games. Not to say that he's not very impressive, or that I don't think it is possible that he could have been better than he is if he'd always put 100%, but I'm just not sure I see that. In other news: Midas + Show Spoiler + Through Month Wins Losses Games Percentage Nal_rA + Show Spoiler + Through Month Wins Losses Games Percentage | ||
|
hypercube
Hungary2735 Posts
One graph that I think would be very interesting is a moving average of win percentage in the last n games (where n is something between 50 and 100). | ||
|
Mumei
United States254 Posts
On February 03 2011 08:03 hypercube wrote: How do you get your data? Do you have a game list that you can analyze automatically or are you searching TLPD by hand? One graph that I think would be very interesting is a moving average of win percentage in the last n games (where n is something between 50 and 100). TLPD, and by hand, so I only do players I'm interested in. I use the standard leagues category, since that is what TL uses for Elo. As for your graph, do you mean something where it calculates, say, Jaedong's average over games 495 -> 595, and then as he plays more games, it calculates average from 496 -> 596 and 497 -> 597, and so forth? I agree that it would be interesting, but it seems like it'd be awful time-consuming to keep it updated manually. | ||
|
GenesisX
Canada4267 Posts
| ||
|
hypercube
Hungary2735 Posts
On February 03 2011 09:19 Mumei wrote: TLPD, and by hand, so I only do players I'm interested in. I use the standard leagues category, since that is what TL uses for Elo. Wow, that looks like a lot of work. As for your graph, do you mean something where it calculates, say, Jaedong's average over games 495 -> 595, and then as he plays more games, it calculates average from 496 -> 596 and 497 -> 597, and so forth? I agree that it would be interesting, but it seems like it'd be awful time-consuming to keep it updated manually. Yeah, that's what I meant. I wonder what the highest 100 game winrate was. It also gets rid of the artificial highs and lows at the start of the players career. And you probably need at least 100 games to make a meaningful distinction between an 80% and 82% winrate. | ||
|
Mumei
United States254 Posts
On February 04 2011 09:39 hypercube wrote: Wow, that looks like a lot of work. Yeah, that's what I meant. I wonder what the highest 100 game winrate was. It also gets rid of the artificial highs and lows at the start of the players career. And you probably need at least 100 games to make a meaningful distinction between an 80% and 82% winrate. Just based on looking at things on TLPD at a glance, it looks like Flash. oov's highest win-rate was at the start of his career, so at 100 games, he had 77% (77 - 23). The same thing is true of Nada, who had 71.29% at 101 games. Flash has gone 235 - 74 since January 2009, or 76.05% over the course of 309 games, so there's obviously a lot of room in there for him to have broken 77%.. He had 82% between October 15, 2009 through May 5, 2010. for instance; he might've gotten something higher, too. Bisu is currently on a 78.55% run in the last 52 games. Jaedong doesn't seem to go on those crazy tears; he seems pretty consistently at ~69% - 72% over extended periods. And yes, Flash probably has the record for ~50 games, too; when Flash hit 2443 Elo, he had won 88% of his previous 50 games. | ||
| ||
