[Spoilers] Effort's record - Page 13
Forum Index > BW General |
Ideas
United States8058 Posts
| ||
MoRe_mInErAls
Canada1210 Posts
On June 16 2009 11:56 Ideas wrote: bare in mind ELO inflation. if there was some sorta ELO inflation I tihnk Boxer or Oov would defiantly sitll be #1. Also more difficult to dominate now than it was when Boxer or Oov played, so it evens out. | ||
Avidkeystamper
United States8551 Posts
| ||
Magic84
Russian Federation1381 Posts
On June 16 2009 12:15 Avidkeystamper wrote: No it doesn't. Nowadays players play a lot more proleague games against mediocre players that's basically free ELO points that a player missed out on since they couldn't play as many people when they were godly. As far as i know inflation is tied with skill level. Numbers not just inflate because of time, but because players get better overall. At least it's like that in reality, even if it's not supported by math. Players are better now, and it's tougher to dominate, you need higher skill level and faster hands than before. | ||
![]()
Hot_Bid
Braavos36372 Posts
On June 16 2009 11:14 p4NDemik wrote: When was the last time a player not named Flash, Bisu, or Jaedong was #1 ELO? Did JangBi ever manage to get to #1? It would be so weird for a new player to be #1, and this is coming from a huge EffOrt fan. iirc the last person not jd/flash/bisu to have #1 ELO was stork for a short while not very sure though | ||
Husky
United States3362 Posts
| ||
![]()
Ver
United States2186 Posts
On June 16 2009 11:58 MoRe_mInErAls wrote: Also more difficult to dominate now than it was when Boxer or Oov played, so it evens out. ELO inflation does exist in some form or another. oov's domination is unparalleled by any of these new guys yet they all have higher elo peaks. Also Flash/Jaedong have their ELO peaks in 2009 despite having their best streaks back in 2008 As for the difficulty to dominate that's just a myth for the most part, at least once progaming was solidified in oov's era. The only new difficulty is the much more frequent proleague which over-schedules and exhausts players. Despite progaming having grown much more since oov's time Flash/Jaedong/Bisu still manage to dominate. Only difference is that none of them can eclipse the others, possibly due to the scheduling issues. It's all about skill relative to the age. And Boxer's ELO is never going to be anywhere near what it should be because TLPD (understandably) doesn't have the info from a lot of the major tournaments he played in back then (before ogn/mbc solidified themselves). On June 16 2009 10:36 B1nary wrote: Can you elaborate a bit on what you mean by Leta not having the tools necessary to take down good players? Do you mean strategy, mechanics, mindset? Or something different? A mix. Leta does not take many risks. He tries to overpower his opponent just by outplaying them, which is very good when facing your random proleague guys like Calms but runs into trouble against starleague contenders who get big build order or tactical advantages versus him and run him into the ground. He plays to his strengths and those are his mechanics and harass, not his strategy. | ||
Avidkeystamper
United States8551 Posts
| ||
Shikyo
Finland33997 Posts
On June 16 2009 12:35 Hot_Bid wrote: iirc the last person not jd/flash/bisu to have #1 ELO was stork for a short while not very sure though IIRC that's correct, back when he won his first OSL. That's why I mentioned him. =) | ||
GoSu
Korea (South)1773 Posts
![]() | ||
Nevuk
United States16280 Posts
Edit : I see that we won't know his OSL opponents til Friday, and his next msl matches won't be til after the rest of ro32 finishes, a week or two from now. It's unlikely for him to face any difficult opponents anytime soon, the strongest opponent we know he'll face for sure is Violet, and while I like Violet, it won't say much about Effort if effort wins that. Who would the worst possible OSL ro16 opponent be? Jaedong? Bisu? Flash? Fantasy? All are possible. | ||
Ace
United States16096 Posts
On June 16 2009 12:54 Avidkeystamper wrote: I have to say, you probably get this a lot, but Ver is one of the best SC-related posters on this site. It's not important whether I agree with you, but your posts concisely address the issue and are very clear and organized. it's because he's Mafia | ||
Goragoth
New Zealand1065 Posts
On June 17 2009 05:39 Nevuk wrote: Who would the worst possible OSL ro16 opponent be? Jaedong? Bisu? Flash? Fantasy? All are possible. Without a doubt Jaedong. A quick look at his record shows that he has only lost one BoX ZvZ (against July at WCG Korea 2007) and his ZvZ has only improved since then. Sure, EffOrt took a game off him recently but knowing Jaedong that will just make him even more likely to win the next time they meet. Of course Bisu would be very tough as well, and neither FlaSh nor Fantasy are pushovers, but Jaedong has to be the one person he would absolutely not want to face. | ||
qrs
United States3637 Posts
On June 16 2009 12:54 Avidkeystamper wrote: I have to say, you probably get this a lot, but Ver is one of the best SC-related posters on this site. It's not important whether I agree with you, but your posts concisely address the issue and are very clear and organized. agreed; whenever I see Ver's name attached to a post, I make a point of reading it. | ||
cz
United States3249 Posts
On June 16 2009 12:15 Avidkeystamper wrote: No it doesn't. Nowadays players play a lot more proleague games against mediocre players that's basically free ELO points that a player missed out on since they couldn't play as many people when they were godly. That's not how ELO works. There is an argument to be made that a players ELO was never representative of his skill due to not a large enough sample (due to the scarcity of games in pre-proleague games), but there are no 'free' or 'easy' or even 'easier' points in ELO. Your expected change in your ELO rating in a game vs a 2300 or 2000 player is the same, zero, as long as: 1) You (and your opponent) have played a large enough sample of games, respectively. 2) Your skill level or your opponent's skill level hasn't changed recently. Inflation can occur when a lot of new players play a few games and don't play again, and their skill is under the 2000 start point. | ||
Avidkeystamper
United States8551 Posts
ELO is not a perfect representation. That's why players go on hot streaks where they maintain a winning % that is nowhere near their overall record, thus allowing them to reach their peak. How many games they can play will be directly proportional to how high their peak is. | ||
cz
United States3249 Posts
On June 17 2009 08:00 Avidkeystamper wrote: Yes, but the evidence is to the contrary. Sure if you play an infinite amount of games with a constant skill level, your ELO will stay constant, but can we really say that for someone like Boxer, his ELO is only a measely 225x? He mostly played in individual tournaments where the level of skill was higher than the proleague (comparatively). So there is inflation, and it is "free ELO points". I mean Effort's getting a lot more opportunities to play during this hot streak of his that other older players may not be able to benefit from due to a dearth of games. As long as the two conditions I mentioned are met, there are no free points or easy points. All points are equally difficult to get. I also think you don't understand the basis of inflation in an ELO system: it doesn't come from everyone playing more games, it comes from new, overrated players (they are given 2000 ELO to start) playing a few games and then not playing again. If you consider playing vs those players 'easy points', then you are correct, as they don't satisfy the two criteria I posted about previously. As to your edit, ELO has nothing to do with winning % or winning streaks, etc. There is naturally variance in starcraft, and so you will see at some points that ELO is not precisely representative of a players skill. Once the player plays enough games, though, that evens out. | ||
Avidkeystamper
United States8551 Posts
| ||
cz
United States3249 Posts
On June 17 2009 08:05 Avidkeystamper wrote: I don't think we're talking about the same "inflation". A general trend whereby the top players' (whoever they may be) rating increases over time? | ||
cz
United States3249 Posts
| ||
| ||