[sondr3]

Which is what I like about Dota, with LoL it's usually blatantly obvious who will win by the 10 minute mark but it's not so much so with Dota. Obviously stomps happen and the team that leads at 10 more often than not wins but it doesn't feel as loop sided as LoL does.

[tomsthumb]

> with LoL it's usually blatantly obvious who will win by the 10 minute mark

maybe in the pro games, but even then that's a pretty iffy statement. a bad engage in a teamfight or surprise gank can swing things pretty hard if the teams know what they're doing. in anything below, say, high diamond there's just too many mistakes to make this claim.

[jamesaguilar]

I have heard this view, but is it actually supported by the data? That is, do comebacks happen with any significant frequency in high-level DotA?

For example, in all of the finals of the recent International, there was a leader at 10m who won the game. I'm not going to go through all the games and check, but I'd guess we'd find this to be true the vast majority of the time in professional DotA games, as it is in League games.

Not to make too sweeping of a generalization, but I expect it is inherent in games with an economy and relatively low random factors. On top of the normal expectation that who gains a lead is likely to be more skilled, economies that allow you to build additional strength via the dominance conferred by your current strength reinforce existing leads.

[chc]

AFAIK the "deathball" strategy recently became really dominant leading up to The International. The most recent balance patch after The International included a lot of measures to swing it back the other direction.

[NamTaf]

The most recent patch introduced a comeback mechanic that was so powerful that it resulted in comebacks becoming the norm. It got toned down within 2 weeks.

Comebacks are fairly frequent in dota.