AIs have unfair advantages

[zookeeper]

yes, but 70% chance to hit and 3 attacks means you will miss 1 and connect the other 2.

No, it doesn't. It means you have a 70% change to hit on each attack.

[deoxy]

yes, but 70% chance to hit and 3 attacks means you will miss 1 and connect the other 2.

No, it doesn't. It means you have a 70% change to hit on each attack.

Yes, it does mean that, but only on average, over a large number of attempts.

You want odds? Here they are:

3 attacks, 30% hit chance:

0 hits: 0.343
1 hit: 0.441
2 hits: 0.189
3 hits: 0.027

so, getting all three hits should be about 1 time in 37... not 37 GAMES, 37 attacks. So, if you've got an average of 3 such guys on the board, and they each engage in combat an average of 1 per round (including attacks from both sides), then you'll see that happen about every 12 turns.

So, it's not INCREDIBLY rare...

I've attacked a 70% defense unit 18 times in one turn, with only one hit. The odds on that one are pretty phenomenal... but it's only happened once.

PS: just because I'm ridiculously anal, I made an Excel spreadsheet to calculate the odds for you. I'll attach it to this post. Just put the number of attacks in B2, and then find your hit-chance column. It does the rest, and it handles up to 10 attacks. If you want AT LEAST a certain number of hits (say, hitting at least twice), add up the numbers for that number of hits and greater (so, with three attacks, you would add the 2 hits row and the 3 hits row together); if you want AT MOST a certain number of hits (say 1), you add the numbers for that number of hits and fewer (so, with three attacks, you would add the 0 hits row and the 1 hit row together).

[Radaga]

No, it doesn't. It means you have a 70% change to hit on each attack.

Yes, it does mean that, but only on average, over a large number of attempts.

You want odds? Here they are:

3 attacks, 30% hit chance:

0 hits: 0.343
1 hit: 0.441
2 hits: 0.189
3 hits: 0.027

so, getting all three hits should be about 1 time in 37... not 37 GAMES, 37 attacks. So, if you've got an average of 3 such guys on the board, and they each engage in combat an average of 1 per round (including attacks from both sides), then you'll see that happen about every 12 turns.

0.027 is 2,7% in 100 attacks, 2, maybe 3 will hit fully.

Lets put a few elves at wood tiles, surounded by 10 neutral enemies with 3 ttacks each, and let them attack.

In 10 rounds, only 1 of them will hit thrice.

Or will it?

[Blarumyrran]

sometimes it seems like the RNG is evil and biased against you, i know it, recently i had got an elder wose miss 8 hits in a row against a poacher, causing it to gain enough xp to level, with him having 60% defence (thats 0,6^8= about 1,7 %). but the effect is rather psychological, you just dont remember hitting many times accurately against the chances as much as missing. so let us not be emo and let us rather rejoice over every hit that we have not missed.

oh and the most convincing argument i know of, Blackjack is like totally 99% random unless cheating! pokers very random too, to a lesser extent. and still they are so much played. so let us not be emo and let us rather rejoice that wesnoth is not even MORE random than it currently is.

[Clonkinator]

Why do so many guys freak out if they have a CTH of 60% but hit only, say, 59% of the time? The AI can't cheat. Hey, in SX I once had a Great Mage upgraded to 45 ranged attacks (yes, magical!). However, it actually managed one time to miss every single of these 45 shots. The funniest thing was, however, after that the statistics window STILL showed me I had done a lot more damage than expected. But hey, take it easy. This is a game, not real life!

[Radaga]

the thing that concerns me most is not cheating, but imperfect algorythims, anyway.

But yeah, it is a game... or is it ?