EPT[HomeStory Cup XX] Championship Sunday - Page 20
| Forum Index > StarCraft 2 Tournaments |
|
Zambrah
United States7393 Posts
| ||
|
sneakyfox
8216 Posts
On November 25 2019 02:15 Penev wrote: Inno might as well wear an eye patch at this point, good for both pirate and patch terran memeness as well That would be great but with the machine running like this I think it will be difficult to make the patch terran memes :| | ||
|
sirok_
33 Posts
On November 25 2019 02:17 ZigguratOfUr wrote: Special is overstating things a bit. The map pool is unfavourable for terran in TvZ but not unduly so: + Show Spoiler [All Liquipedia Results] + ![[image loading]](https://i.imgur.com/LgPE4TP.png) + Show Spoiler [Filtered to only include top 100 Aligu…] + ![[image loading]](https://i.imgur.com/efx6QZ3.png) In a system where players get matched according to their ELO / prestige, win rate does not mean much. I believe it would made much more "mathematical" sense to assume top 5 Z should approximately equal top 5 T and compare the win rate at this level. | ||
|
TheAnarchy
Chile1105 Posts
On November 25 2019 02:16 sirok_ wrote: Why no new rising stars among Korean would imply that the current best Koreans players are not practicing ? Even if that argument made sense - it does not - the same logic could apply for top EU. Who are the rising stars among top EU players ? Serral and Reynor are both playing this game since WoL too. Koreans are not training in general. So new blood isn’t rising because they don’t train. Is not a thing of talent only. Old players have already developed their potential so they don’t need to train as hard to mantain close lvl, that’s a fact in most sports and more in the ones that are more intelectual. NA always trained less then Koreans they didn’t had proteams, so they maintained the lvl of training. In WOL there other foreigners in their peak. Snute, scarlet, neeb, etc. Serral and Raynor are new compared too the old foreigners. | ||
|
ZigguratOfUr
Iraq16955 Posts
On November 25 2019 02:20 sirok_ wrote: In a system where players get matched according to their ELO / prestige, win rate does not mean much. I believe it would made much more "mathematical" sense to assume top 5 Z should approximately equal top 5 T and compare the win rate at this level. If you only consider the top 10 players then your sample size is non-existent and your results have no statistical significance. Additionally there is a clear difference between different maps, so this at least informs about how good maps are relative to each other. | ||
|
sirok_
33 Posts
On November 25 2019 02:22 ZigguratOfUr wrote: If you only consider the top 10 players then your sample size is non-existent and your results have no statistical significance. Additionally there is a clear difference between different maps, so this at least informs about how good maps are relative to each other. I don't deny a sample size of 10 sucks from a statistical standpoint. I still think it's better than using the result of a top 3 Terran vs Top 20 Zerg as an indicator of map balance. | ||
|
deacon.frost
Czech Republic12129 Posts
| ||
|
sneakyfox
8216 Posts
| ||
|
ZigguratOfUr
Iraq16955 Posts
On November 25 2019 02:25 sirok_ wrote: I don't deny a sample size of 10 sucks from a statistical standpoint. I still think it's better than using the result of a top 3 Terran vs Top 20 Zerg as an indicator of map balance. The top 3 terrans just don't play enough tournament games on each map for that. Besides your concern over elo unevenness in matchups applies equally for different maps, so it shouldn't bias the results. | ||
|
sneakyfox
8216 Posts
| ||
|
Penev
28524 Posts
| ||
|
sneakyfox
8216 Posts
| ||
|
Durnuu
13322 Posts
| ||
|
SetGuitarsToKill
Canada28396 Posts
| ||
|
JJH777
United States4416 Posts
| ||
|
deacon.frost
Czech Republic12129 Posts
| ||
|
sudete
Singapore3054 Posts
Totally outskilled | ||
|
parksonsc
175 Posts
| ||
|
Penev
28524 Posts
| ||
|
AzAlexZ
Australia3303 Posts
| ||
| ||
