EPTGeneral Discussion - Page 540
| Forum Index > Dota 2 General |
IRC chatter should remain in IRC - http://webchat.quakenet.org/?channels=tl.dota2 Posts that relate to topics with their own thread, such as in News, Tournaments or Strategy should go in those threads. | ||
Firebolt145
Lalalaland34503 Posts
| ||
TheYango
United States47024 Posts
Whether you consider this a systemic design issue that supports are not well suited to carry low-level pub games is up to you. I'm fairly certain that most do not. It's regrettable that a game that is designed around 5v5 play cannot be balanced around solo play as well, but it's simply not worth potentially jeopardizing balance in organized 5v5 play (which is the intended mode of play for competitive games--not solo pub play) for the good of pubs. Essentially, it's the exact same sacrifice Blizzard makes not balancing SC2 around 2v2, 3v3, or 4v4. You can't achieve balance in both solo and team-play, so you have to make sacrifices for the mode which you feel should be the primary mode for competitive play. | ||
|
Skullflower
United States3779 Posts
http://www.joindota.com/en/news/3073-admiration-and-vigoss-reunite-in-moscow5 | ||
|
MrCon
France29748 Posts
On May 12 2012 22:24 paralleluniverse wrote: As I said, it's not just Riki and Ursa, they were examples of carries. It's carries in general. If carries were OP, the optimal strategy would be to pick 5 carries. | ||
|
JeeJee
Canada5652 Posts
So yes, those stats show pretty well that certain heroes (e.g. invoker) are hard to succeed with as a beginner, and certain heroes (e.g. ursa) are easy. Comparing heroes to units in SC2 is a wrong way to go about it. You should be looking at strategies, e.g. 4gate. I don't play SC2 but I suspect 4gate is like ursa -- easy to do, so very strong at low level of play. But, with a skill cap, so I suspect tournament level games don't see 4gate win all day. Is that true? I'm genuinely curious. Either way, I think someone should scrape high level, tournament games instead. If MMR gets publicly visible, maybe add in top MMR games in there as well, for a larger sample size. | ||
|
BeanerBurrito
1010 Posts
Ursa didn't suddenly become op in the last few weeks, when he wasn't for the last few years. If you are making a claim that a hero is op because that hero is good in pubs, you are missing the point of balance. Balance does not make it so that people of any skill level are only able to play at the same level, that is dumbing the game down. In a pub if you are unable to coordinate your team to kill the ursa, the problem isn't that ursa is op, the problem is that your team isn't coordinating and/or your skill level isn't high enough to beat your opponents. Your skill level should be completely separate from the balance of the game, meaning that even if ursa is able to kill a whole bunch of low skilled players, it does not mean anything in relation to whether he is balanced or not. Balance is about the capability of each team of heroes vs the other team, if your team was capable of killing ursa but unable, the game is balanced regardless of whether or not your whole team died to said ursa. | ||
|
Ack1027
United States7873 Posts
| ||
CountChocula
Canada2068 Posts
On May 12 2012 22:37 paralleluniverse wrote: If you have evidence of bias please show it. Otherwise, stop the baseless rumormongering. Yeah rabidch, you rumour-mongerer, you. | ||
|
Kipsate
Netherlands45349 Posts
| ||
|
Dead9
United States4725 Posts
On May 12 2012 22:37 paralleluniverse wrote: If you have evidence of bias please show it. Otherwise, stop the baseless rumormongering. lol + Show Spoiler + bi·as [bahy-uhs] 1. an oblique or diagonal line of direction, especially across a woven fabric. 2. a particular tendency or inclination, especially one that prevents unprejudiced consideration of a question; prejudice. 3. a systematic as opposed to a random distortion of a statistic as a result of sampling procedure. | ||
|
cilinder007
Slovenia7251 Posts
| ||
|
Coal
Sweden1535 Posts
On May 13 2012 04:23 JeeJee wrote: I think statistical analysis of heroes and matchups is a good idea, but scraping all public games for info is a wrong way to go about it. Certain heroes scale better with skill, so by definition they will either be overpowered or underpowered at low/high skill levels. You can't balance e.g. invoker for both low level and high level play. At this point you have to decide which is more important. Statistics of all games will naturally show you results skewed toward lower level games, simply because there are more of them. So yes, those stats show pretty well that certain heroes (e.g. invoker) are hard to succeed with as a beginner, and certain heroes (e.g. ursa) are easy. Comparing heroes to units in SC2 is a wrong way to go about it. You should be looking at strategies, e.g. 4gate. I don't play SC2 but I suspect 4gate is like ursa -- easy to do, so very strong at low level of play. But, with a skill cap, so I suspect tournament level games don't see 4gate win all day. Is that true? I'm genuinely curious. Either way, I think someone should scrape high level, tournament games instead. If MMR gets publicly visible, maybe add in top MMR games in there as well, for a larger sample size. Pretty much, tho 4gate isn't nearly as strong as it used to be ^_^ When it was popular even high (like really high) protoss players used tho.. | ||
|
Benjef
United Kingdom6921 Posts
Like is Blue guy top MMR? | ||
|
Hoban
United States1600 Posts
On May 13 2012 06:58 Benjef wrote: So I que for all the game modes and every so often I get CM I'm almost always captain does that mean I have the highest MMR on my team or I'm just one lucky guy? Like is Blue guy top MMR? Pretty sure it determines position by MMR so yes I would assume you generally have the highest MMR in your CM games. | ||
|
Erasme
Bahamas15899 Posts
| ||
|
NB
Netherlands12045 Posts
| ||
|
Stancel
Singapore15360 Posts
i don't know wtf is going on but i love it | ||
|
paralleluniverse
4065 Posts
On May 12 2012 23:53 Chiharu Harukaze wrote: Statistics don't mean anything if they have no practical interpretation. Your samples sizes are small, and there is far too much noise in the data due to the incomplete matchmaking system as well as other extenuating factors. That's not even bringing in the issue of correlation not implying causation either. Stop making shit up. You don't know a damn thing about statistics. How do you know it has no practical interpretation? Where is your source that the sample size is small? How do you know there is too much noise in the data? Do you know how to read error bars? There is clearly a trend of >50% win rates from the data, even when the error bars are small, and when they're large they the imbalance could be even worse than it is reported, There is a functioning matchmaking system, which attempts to move everyone's win rates to 50%. And these win rates don't even control for skill level like Blizzard's. Blizzard's unadjusted win rates are all 50%. It's only after adjusting for skill level they are between usually 50-55%. The fact that even if a matchmaker that moves everyone's win rates to 50%, they still aren't 50%, is a colossal fuck up. Balance also isn't a causation issue. A causation issue would be whether or not Ursa's overpowered spell makes him overpowered. Balance is by definition whether or not a win rate close to 50%. The stats show that Ursa is imbalanced, by definition. It doesn't show what causes Ursa to imbalance (too much damage, too much HP, attacks too fast, etc), but it shows that he is. | ||
|
paralleluniverse
4065 Posts
On May 13 2012 02:17 superstartran wrote: 1) Your statistics are for one taken from public matchmaking games, therefore invalidating the discussion already. Various heroes like Invoker are extremely difficult for low level players for play, and yet are highly banned all the time because he is a fucking team fight monster that does massive AoE and has huge amounts of disables at his disposal ranging from Cold Snap, Tornado, Deafening Blast, and Ice Wall. Heroes like Lich and Ursa are effective particularly at low levels of play because they are so braindead easy to play, and are effective because people don't know how to play against them, not because they are good at that level of play. 2) You're a fucking idiot if you think Ursa is any way shape or form overpowered. He is the ultimate pub stomp hero because public players do not buy wards, do not ever gank jungle, and always let Ursa get ahead on levels/farm by taking Roshan early. He is such an easy hero to counter even with all the farm in the world, since he is god awful easy to kite, and is smashed by superior stuns like Beastmaster Roar, Bane Elemental's Ult, and Batrider lasso. 3) You're obviously bad as fuck, or a huge fucking troll. Stop polluting this forum because it's blatantly obvious you have no idea what you're talking about. We've already listed the counters and why your graphs are simply wrong. If you take a sample of competitive games, you'll find that certain heroes are picked/banned far more than Ursa for a damn reason. I could go on and on about how you're wrong, but it's not worth my time. The only graphs in Starcraft 2 that are noteworthy in the first place are the ones that involve only professional level of play. The graph you just listed is the equivalent of attempting to utilize Battle.Net winrates to justify a balance claim, which basically means you're either a huge idiot, or a massive troll. Where are these mythical high level stats? User was temp banned for this post. | ||
|
NB
Netherlands12045 Posts
On May 13 2012 13:21 paralleluniverse wrote: Stop making shit up. You don't know a damn thing about statistics. How do you know it has no practical interpretation? Where is your source that the sample size is small? How do you know there is too much noise in the data? There is a functioning matchmaking system, which moves everyone win rates to 50%. And these win rates don't even control for skill level like Blizzard's. Blizzard's unadjusted win rate are all 50%. It's only after adjusting for skill level they are between usually 50-55%. The fact that even if a matchmaker that moves everyone's win rates to 50%, they still aren't 50%, is a colossal fuck up. Balance also isn't a causation issue. A causation issue would be whether or not Ursa's overpowered spell makes him overpowered. Balance by definition whether or not a win rate close to 50%. The stats show that Ursa is imbalanced, by definition. It doesn't show that attributable of Ursa causes the imbalance (too much damage, too much HP, attacks too fast, etc). so tell me 1 thing: is 4 gates considered imbalance in sc2? bc majority of PvX games played in master league that the Protoss executed 4 gates, he won. This is true and the statistic was taken soon after sc2 was first released. If it was imbalance, why Protoss never win a single major tournaments in the first year of sc2 except MC? There is a requirement that to be met in EVERY balance debate: You need to play the game at the top level. Are you playing ANY competitive game at top level? You really think that your stats would mean anything if you dont even understand the game? All you brought up in this thread was stat 1, stat 2. How about you pick ursa riki in 1 team and im gona be on the other side, show me how the hero is really imbalance? On another topic: is match history only limited to 72 pages? im pretty sure i have played way more games than 7x72 -_- | ||
| ||
