I wasn't talking about that level. Imagine a starting setup (or better yet, a series of setups). And an ideal level of development, say, in 2 weeks. Let's consider that as 100%.
Then, we give this setup to different players and see what percentage of the ideal they achieve (in a real game, not offline). This is the famous level of effectiveness. We average it across the setups.
I've analyzed many saves from different players, and I have a pretty good understanding of the level of play. So, I often see a lot of mistakes in top players in SoD that are not typical for top players in TE.
Recently, I watched Vitalik's game against Sirus, which he posted in the interportal thread. He was afraid to attack the lizardmen horde on the grass with his flux army (with water elementals), a hero with the "Feast" ability, and 7 basilisks from Bron. I, on the other hand, broke them with the loss of 2 air units and 2 water elementals, without using magic (and it would have been even easier with a spellbook and haste). I've broken them many times in real games, too, and without basilisks (for example, playing with Invisibility).
And this is a player who has a tournament win rate of around 90% in SoD. I still have a long way to go...
Wicc, I'm replying in this thread because otherwise it will be off-topic.
I don't quite agree, and here's why. There's the result of vitalic in the Portal match against Bay. To reiterate, the result was 2-0 in favor of vitalic. Therefore, I don't find this example very convincing. There are good players on Bay, no one disputes that. But playing as a team (!!!) with morale disabled for neutral creatures or in a real game is a big difference. You can take almost any online match and, by analyzing the saves, find opportunities for improvement, etc. The higher the element of randomness (and it's higher in SoD than in other versions of the game), the greater the risk, and a player's skill in SoD also includes assessing game risks...
P.S. The article is good.
