EPTThis is the on going tournament btw,
PSL/2Aligulac.com changelog and feedback thread - Page 34
| Forum Index > SC2 General |
|
DLSMIZEL
Philippines16 Posts
This is the on going tournament btw, PSL/2 | ||
|
pewy
2 Posts
Pink Drip Pylon/Seeding | ||
|
pewy
2 Posts
Pink Drip Pylon/Seeding | ||
|
chickenandbeer1234
5 Posts
I just saw Maru's overall rating in every vs race suddenly dropped after vs Clem in team league First of all, before he goes vs Rogue in GSL final, Maru's rating was 3071 vP, 3043 vT, 3213 vZ on LIST 292: May 5th And now, Maru's overall rating dropped significantly by 3023 vP, 3023 vT, 3136 vZ on LIST 293: May 19th for some reason The thing I cannot understand about this is that Maru only played two games between May 5th and May 17th which are vs Rogue in GSL and vs Clem in Team league. Here are my 2 questions 1. Why Maru's rating vs protoss suddenly dropped when he didn't even play a single game vs protoss between May 5th and May 17th? 2. Clem had the highest rating even vT (3087) on May 5th according to aligulac, and even when Clem and Maru tied 1-1 Clem's rating in vT even went higher to 3113, and meanwhile Maru's vT rating when lower to 3023 from 3043. This looks very weird since a player rated with 3087 and another player rated with 3043 played each other and tied as 1-1, but 3087 -> 3113, 3043 -> 3023? I also compared with both Maru and Clem's every TvT match ups between May 5th and May 17th. And it looks like Maru only played one TvT, which is vs Clem yesterday between May 5th and today. But his TvT rate was dropped even when he tied vs higher rated TvT player. And meanwhile, Clem's result in TvT between May 5th and May 19th seems much worse as 1-3 Cure, 2-3 Cure, 0-1 souL, 3-0 ByuN (ByuN only has 2800s rating in TvT according to aligulac), and several wins vs much lower rated foreigners. And despite all of this result, Clem vT rating actually went higher to 3113(May 19th) from 3087(May 5th). And Maru's vT rating fell from 3043 to 3023 after he tied 1-1 with 3100's rated vT player. I love both Clem and Maru they are my favorite players, but I must point out about this rating system that is not understandable. I felt strange about aligulac a lot even before today. But even though we know Aligulac is with the 'Automatic rating system', we should also not forget it can be manipulated by 'A person' or 'People' who want to overhype their top players by changing the ratings without noticing others. I believe this is possible. | ||
|
GoloSC2
713 Posts
On May 17 2021 16:26 chickenandbeer1234 wrote: ... Following you post I looked a bit into the FAQ, and this stood out to me as to why Maru's TvP might've decreased even though he didn't play the matchup (emphasis mine): Rating adjustments The rating adjustments can seem a bit off. Why is that? Common complaints include players gaining or losing points when they score exactly as expected, or gaining points when they underperform, and losing points when they overperform. Several things can impact this. First of all, the expected scores are rounded to one decimal place, so you're not seeing the whole story. (An expected score of 1.0–0.0 could in reality be 0.96–0.04 for example.) Also, you should keep in mind that the rating adjustments for each of the four categories (general rating, and each matchup rating) are not done independently of each other. The system will try to update each rating according to the certainty bias (mentioned earlier), but it's also restricted by the fact that the matchup ratings must have mean equal to the general rating. The upshot of this is that if a player overperforms versus Terran (say 2–0 when 1–1 was expected), but significantly underperforms in the other matchups (say 0–10 in each when 5–5 was expected), the rating versus Terran may still decrease. His match vs Rogue was predicted to be in his favor (3.1 to 1.9) and he ended up loosing 1-4, so this loss is probably the reason for the overall decline. Now you could argue that this doesn't make a whole lot of sense, in which case I'd urge you to look through the FAQ and familiarize yourself with the methodology. | ||
|
Gottingen
United States27 Posts
| ||
|
Luepert
United States1934 Posts
| ||
|
nephest0x
1 Post
| ||
|
neptunusfisk
2286 Posts
On May 17 2021 16:26 chickenandbeer1234 wrote: I felt strange about aligulac a lot even before today. But even though we know Aligulac is with the 'Automatic rating system', we should also not forget it can be manipulated by 'A person' or 'People' who want to overhype their top players by changing the ratings without noticing others. I believe this is possible. Maru played vs rogue and lost badly, his overall rating dropped (-48) He had no matches vs protoss, so his vP rating dropped the same as his overall rating (-48) He played vs clem and did better, so he lost less rating vT (-20) Also, see the FAQ: Several things can impact this. First of all, the expected scores are rounded to one decimal place, so you're not seeing the whole story. (An expected score of 1.0–0.0 could in reality be 0.96–0.04 for example.) Also, you should keep in mind that the rating adjustments for each of the four categories (general rating, and each matchup rating) are not done independently of each other. The system will try to update each rating according to the certainty bias (mentioned earlier), but it's also restricted by the fact that the matchup ratings must have mean equal to the general rating. The upshot of this is that if a player overperforms versus Terran (say 2–0 when 1–1 was expected), but significantly underperforms in the other matchups (say 0–10 in each when 5–5 was expected), the rating versus Terran may still decrease. | ||
|
TheBB
Switzerland5133 Posts
My apologies if I missed it being documented somewhere, but is there a recommended query timeout when using the API? There's no hard limit. I haven't seen any undue load on the server, so I assume whatever you're currently doing is fine. It also depends a bit on whether the query you're running is something expensive. If possible, for high load API usage I recommend downloading the database dump whenever it is updated, loading it in a local database and then making queries there instead. The data is the same anyway. And you can probably do more exotic things. Is there a way to revoke the API key? Please e-mail me the trash key, I'll remove it. | ||
|
dbRic1203
Germany2655 Posts
On May 17 2021 16:26 chickenandbeer1234 wrote: Can someone explain to me about this? I just saw Maru's overall rating in every vs race suddenly dropped after vs Clem in team league First of all, before he goes vs Rogue in GSL final, Maru's rating was 3071 vP, 3043 vT, 3213 vZ on LIST 292: May 5th And now, Maru's overall rating dropped significantly by 3023 vP, 3023 vT, 3136 vZ on LIST 293: May 19th for some reason The thing I cannot understand about this is that Maru only played two games between May 5th and May 17th which are vs Rogue in GSL and vs Clem in Team league. Here are my 2 questions 1. Why Maru's rating vs protoss suddenly dropped when he didn't even play a single game vs protoss between May 5th and May 17th? 2. Clem had the highest rating even vT (3087) on May 5th according to aligulac, and even when Clem and Maru tied 1-1 Clem's rating in vT even went higher to 3113, and meanwhile Maru's vT rating when lower to 3023 from 3043. This looks very weird since a player rated with 3087 and another player rated with 3043 played each other and tied as 1-1, but 3087 -> 3113, 3043 -> 3023? I also compared with both Maru and Clem's every TvT match ups between May 5th and May 17th. And it looks like Maru only played one TvT, which is vs Clem yesterday between May 5th and today. But his TvT rate was dropped even when he tied vs higher rated TvT player. And meanwhile, Clem's result in TvT between May 5th and May 19th seems much worse as 1-3 Cure, 2-3 Cure, 0-1 souL, 3-0 ByuN (ByuN only has 2800s rating in TvT according to aligulac), and several wins vs much lower rated foreigners. And despite all of this result, Clem vT rating actually went higher to 3113(May 19th) from 3087(May 5th). And Maru's vT rating fell from 3043 to 3023 after he tied 1-1 with 3100's rated vT player. I love both Clem and Maru they are my favorite players, but I must point out about this rating system that is not understandable. I felt strange about aligulac a lot even before today. But even though we know Aligulac is with the 'Automatic rating system', we should also not forget it can be manipulated by 'A person' or 'People' who want to overhype their top players by changing the ratings without noticing others. I believe this is possible. I like how this Person is completly delusional about how the Alligulac System works. Look into how it actually works, it s all in the page in the FAQ, as pointed out before. Ratings are allways also affected by the other matchups as well. And as you pointed out, it s an automated Rating, so it isn t possible to manipulative the Rating alone, without putting Fake matches in | ||
|
Luepert
United States1934 Posts
http://aligulac.com/results/events/38419/ protosito vs illusionclock Gamergy Season 2 illusionclock won 2-1http://aligulac.com/results/events/50909/ myorg vs jauhis Starcraft2.fi 5V. Cup jauhis won 2-0 | ||
|
andertalets77
144 Posts
| ||
|
uselless
93 Posts
 My deepest respect and appreciation for everyone who has ever worked on or contributed to the project. I know I might be overreacting, but it's been almost too influential | ||
|
Kashim
Poland1236 Posts
On August 28 2021 11:27 uselless wrote: Maybe this is a bit unreasonable, but Aligulac hasn't been updated and we're two days into the new cycle. The last one was the 300th one, so maybe it's being discontinued? Either way, it's been the second best thing Starcraft II has had My deepest respect and appreciation for everyone who has ever worked on or contributed to the project. I know I might be overreacting, but it's been almost too influential don't worry | ||
|
uselless
93 Posts
Glad to hear it :D | ||
|
Paljas
Germany6926 Posts
| ||
|
drcassidyferrercount
2 Posts
| ||
|
Luepert
United States1934 Posts
I've tried order_by=date,id, which gives an error and order_by=date&order_by=id but this doesn't actually give results which are ordered by id. Let me know if there is some way to do this, thanks! edit: ahh! after an hour debugging, I figured it out right after I finally posted to ask a question. I had a typo in my query with orderby instead of order_by so that's why it wasn't working as I expected. | ||
|
PremoBeats
582 Posts
If I understand correctly, it measures the gap of rating points you have if you are in the top 7 to the the rating of the player who is ranked 7. So it doesn't include rank, it "only" measures the distance to rank 7 in rating points, correct? And length is the sum of weeks where someone resides in the top 7? Another thing I wanted to ask: What does "Start" on the HoF page mean? Because it isn't the list on which the given player reached rank 7 for the first time and I couldn't deduct a pattern from looking at various player's links in the start category. Thanks a ton! | ||
| ||
