|
On June 18 2014 23:23 Stress wrote:Show nested quote +On June 18 2014 22:49 stuchiu wrote:On June 18 2014 22:40 SatedSC2 wrote:On June 18 2014 22:11 stuchiu wrote:On June 18 2014 20:56 SatedSC2 wrote:
Aligulac on the case!
:D So was KingKong. Next question. Classic and Maru are comparable given that their ratings are primarily based on ProLeague and WCS KR results. KingKong clearly doesn't fit in with that. I hope you're being facetious... I thought you were being facetious too o_o. Somehow I don't think Classic being statistically favored was the reason why people were upset. Honestly, do you guys think Aligulacs rankings/predictor is a good indicator of who will win? Aligulac is extremely useful for viewing the recent match history of players but SC2 is so volatile and depending on the match-up and a few other factors (slumps or who the opponents are) I don't believe it to be a good indicator of who will win. For example, Forgg is ranked #5 and is the highest ranked TvT player. But if you look at his match history since April, against Terran he is 4-4 against MMA, 3-4 against Happy, and has lost to both Innovation and Jjakji. If you use the predictor he is favored against Flash in a BO3 73% to 27% and in a BO5 78% to 22%. Would people honestly believe that Forgg would beat Flash in a BOX series? After seeing Flashs TvT against Bbyong in Shoutcraft I would imagine the odds are against Forgg even though Forgg is ranked #5 overall and Flash is ranked #30 overall. I understand this is one match-up and only one player but I'm using it to argue my case on why Aligulac isn't the best indicator of who should be favored or not favored. In no way am I bashing Aligulac, it is a very well put together site that contains extremely useful information.
It is a good indicator as long as you compare two players who play in the same region.
|
Fiddler's Green42661 Posts
It's ok. I don't take it for a god or anything. As long as you keep in mind what it does and how it does it, it's fine. The biggest problem is the users rather than the database. A lot of people just use aligulac as the truth without really thinking it through.
|
your Country52797 Posts
I knew Classic had won when I saw 4 new bans, 3 of which were balance whining.
Spoilers!!
|
soo wins the whole thing anyway. no whining needed!
|
if protoss wins 3 GSLs in a row then the only person to blame is soo 
|
gg audience for the finals
at least it's not PvP, though?
|
East Gorteau22261 Posts
On June 18 2014 23:23 Stress wrote:Show nested quote +On June 18 2014 22:49 stuchiu wrote:On June 18 2014 22:40 SatedSC2 wrote:On June 18 2014 22:11 stuchiu wrote:So was KingKong. Next question. Classic and Maru are comparable given that their ratings are primarily based on ProLeague and WCS KR results. KingKong clearly doesn't fit in with that. I hope you're being facetious... I thought you were being facetious too o_o. Somehow I don't think Classic being statistically favored was the reason why people were upset. Honestly, do you guys think Aligulacs rankings/predictor is a good indicator of who will win? Aligulac is extremely useful for viewing the recent match history of players but SC2 is so volatile and depending on the match-up and a few other factors (slumps or who the opponents are) I don't believe it to be a good indicator of who will win. For example, Forgg is ranked #5 and is the highest ranked TvT player. But if you look at his match history since April, against Terran he is 4-4 against MMA, 3-4 against Happy, and has lost to both Innovation and Jjakji. If you use the predictor he is favored against Flash in a BO3 73% to 27% and in a BO5 78% to 22%. Would people honestly believe that Forgg would beat Flash in a BOX series? After seeing Flashs TvT against Bbyong in Shoutcraft I would imagine the odds are against Forgg even though Forgg is ranked #5 overall and Flash is ranked #30 overall. I understand this is one match-up and only one player but I'm using it to argue my case on why Aligulac isn't the best indicator of who should be favored or not favored. In no way am I bashing Aligulac, it is a very well put together site that contains extremely useful information.
Aligulac gets predictions right more often than not. Not near 100% accuracy, naturally, but I think it's a better predictor than most would-be experts on TL and Reddit.
|
United States980 Posts
On June 19 2014 00:13 Zealously wrote:Show nested quote +On June 18 2014 23:23 Stress wrote:On June 18 2014 22:49 stuchiu wrote:On June 18 2014 22:40 SatedSC2 wrote:On June 18 2014 22:11 stuchiu wrote:So was KingKong. Next question. Classic and Maru are comparable given that their ratings are primarily based on ProLeague and WCS KR results. KingKong clearly doesn't fit in with that. I hope you're being facetious... I thought you were being facetious too o_o. Somehow I don't think Classic being statistically favored was the reason why people were upset. Honestly, do you guys think Aligulacs rankings/predictor is a good indicator of who will win? Aligulac is extremely useful for viewing the recent match history of players but SC2 is so volatile and depending on the match-up and a few other factors (slumps or who the opponents are) I don't believe it to be a good indicator of who will win. For example, Forgg is ranked #5 and is the highest ranked TvT player. But if you look at his match history since April, against Terran he is 4-4 against MMA, 3-4 against Happy, and has lost to both Innovation and Jjakji. If you use the predictor he is favored against Flash in a BO3 73% to 27% and in a BO5 78% to 22%. Would people honestly believe that Forgg would beat Flash in a BOX series? After seeing Flashs TvT against Bbyong in Shoutcraft I would imagine the odds are against Forgg even though Forgg is ranked #5 overall and Flash is ranked #30 overall. I understand this is one match-up and only one player but I'm using it to argue my case on why Aligulac isn't the best indicator of who should be favored or not favored. In no way am I bashing Aligulac, it is a very well put together site that contains extremely useful information. Aligulac gets predictions right more often than not. Not near 100% accuracy, naturally, but I think it's a better predictor than most would-be experts on TL and Reddit.
Do you know/have a percentage of how accurate it is? It would also be interesting to see how accurate it is for players in the same region (like Proleague) versus players from several regions (such as a travel event like MLG/DH).
|
Aligulac as it is, is a GIANT base of matches. With some clusterization algorithms and a touch of machine learning, vast data could be turned into great predictors. You also need skilled people, that's borderline PhD material right there.
|
East Gorteau22261 Posts
|
On June 19 2014 00:29 Stress wrote:Show nested quote +On June 19 2014 00:13 Zealously wrote:On June 18 2014 23:23 Stress wrote:On June 18 2014 22:49 stuchiu wrote:On June 18 2014 22:40 SatedSC2 wrote:On June 18 2014 22:11 stuchiu wrote:So was KingKong. Next question. Classic and Maru are comparable given that their ratings are primarily based on ProLeague and WCS KR results. KingKong clearly doesn't fit in with that. I hope you're being facetious... I thought you were being facetious too o_o. Somehow I don't think Classic being statistically favored was the reason why people were upset. Honestly, do you guys think Aligulacs rankings/predictor is a good indicator of who will win? Aligulac is extremely useful for viewing the recent match history of players but SC2 is so volatile and depending on the match-up and a few other factors (slumps or who the opponents are) I don't believe it to be a good indicator of who will win. For example, Forgg is ranked #5 and is the highest ranked TvT player. But if you look at his match history since April, against Terran he is 4-4 against MMA, 3-4 against Happy, and has lost to both Innovation and Jjakji. If you use the predictor he is favored against Flash in a BO3 73% to 27% and in a BO5 78% to 22%. Would people honestly believe that Forgg would beat Flash in a BOX series? After seeing Flashs TvT against Bbyong in Shoutcraft I would imagine the odds are against Forgg even though Forgg is ranked #5 overall and Flash is ranked #30 overall. I understand this is one match-up and only one player but I'm using it to argue my case on why Aligulac isn't the best indicator of who should be favored or not favored. In no way am I bashing Aligulac, it is a very well put together site that contains extremely useful information. Aligulac gets predictions right more often than not. Not near 100% accuracy, naturally, but I think it's a better predictor than most would-be experts on TL and Reddit. Do you know/have a percentage of how accurate it is? It would also be interesting to see how accurate it is for players in the same region (like Proleague) versus players from several regions (such as a travel event like MLG/DH).
That's interesting. How many times has Aligulac predicted the outcome correctly? It can be calculated by Aligulac themselves... then they can implement a meta-predictor for their predictor, and... O.o ...
|
That isn't THAT amazing, i have a voting accuracy of 77% right now (with 65% of possible points though -.-), not enough to win liquibet :/
+ Show Spoiler +Ok you caught me, i just wanted to boast a little 
|
Wow its real hard to predict that a protoss will win a PvT bo7 in a code s semifinal. Thank god we got aligulac to tell us that LOL. Talk about stating the obvious and then trying to take credit for it.
|
United States980 Posts
On June 19 2014 00:53 The_Red_Viper wrote:That isn't THAT amazing, i have a voting accuracy of 77% right now (with 65% of possible points though -.-), not enough to win liquibet :/ + Show Spoiler +Ok you caught me, i just wanted to boast a little
What is the accuracy percentage of the top people that participate in liquibets?
|
On June 19 2014 01:03 johnbongham wrote:
Wow its real hard to predict that a protoss will win a PvT bo7 in a code s semifinal. Thank god we got aligulac to tell us that LOL. Talk about stating the obvious and then trying to take credit for it.
"statistically Classic was favoured against Maru" Nothing wrong with that. Try to be able to take a joke
Edit:
On June 19 2014 01:11 Stress wrote:Show nested quote +On June 19 2014 00:53 The_Red_Viper wrote:That isn't THAT amazing, i have a voting accuracy of 77% right now (with 65% of possible points though -.-), not enough to win liquibet :/ + Show Spoiler +Ok you caught me, i just wanted to boast a little What is the accuracy percentage of the top people that participate in liquibets?
Mevious has made 68% of the available points (I'm on 63%, so also within aligulac's range  )
|
I like Classic, but he really wasn't supposed to win this. This season of Code S lost most of its appeal with Maru out.
|
United States980 Posts
On June 19 2014 01:12 Penev wrote:Show nested quote +On June 19 2014 01:03 johnbongham wrote:
Wow its real hard to predict that a protoss will win a PvT bo7 in a code s semifinal. Thank god we got aligulac to tell us that LOL. Talk about stating the obvious and then trying to take credit for it. "statistically Classic was favoured against Maru" Nothing wrong with that. Try to be able to take a joke Edit: Show nested quote +On June 19 2014 01:11 Stress wrote:On June 19 2014 00:53 The_Red_Viper wrote:That isn't THAT amazing, i have a voting accuracy of 77% right now (with 65% of possible points though -.-), not enough to win liquibet :/ + Show Spoiler +Ok you caught me, i just wanted to boast a little What is the accuracy percentage of the top people that participate in liquibets? Mevious has made 68% of the available points (I'm on 63%, so also within aligulac's range )
That's interesting. I'm curious, do you make your "betting" decisions based on Aligulac? Or do you go from memory/intuition on who you think is better?
|
Mostly memory but if I've missed a lot or I'm not sure I'll check past matches and or aligulac. It's a handy tool but you have to know how to interpret the data
|
On June 19 2014 01:21 Stress wrote:Show nested quote +On June 19 2014 01:12 Penev wrote:On June 19 2014 01:03 johnbongham wrote:
Wow its real hard to predict that a protoss will win a PvT bo7 in a code s semifinal. Thank god we got aligulac to tell us that LOL. Talk about stating the obvious and then trying to take credit for it. "statistically Classic was favoured against Maru" Nothing wrong with that. Try to be able to take a joke Edit: On June 19 2014 01:11 Stress wrote:On June 19 2014 00:53 The_Red_Viper wrote:That isn't THAT amazing, i have a voting accuracy of 77% right now (with 65% of possible points though -.-), not enough to win liquibet :/ + Show Spoiler +Ok you caught me, i just wanted to boast a little What is the accuracy percentage of the top people that participate in liquibets? Mevious has made 68% of the available points (I'm on 63%, so also within aligulac's range ) That's interesting. I'm curious, do you make your "betting" decisions based on Aligulac? Or do you go from memory/intuition on who you think is better?
I'm at 72% correct voting percentage, entirely tuition. Where would the fun be using aligulac, especially if that's what a large number of people do?
|
United Kingdom31935 Posts
Maru lost? That kinda sucks
|
|
|
|
|
|
EPT
