Yeah I read it, but isn't he only talking about the new difficulties?
He is BUT his comment doesn't mean in any way shape or form that the other AI's don't cheat. In fact, the comment "We all know that both the new AI's get resource bonuses" is actually referring to the fact that we all already knew the other AI's (Hard and Unfair) cheated via resource bonuses, so it's only obvious that the new even harder AI's (Cruel and Vicious) would also cheat via resource bonuses to a much greater degree, and then he included his observations as to the additional cheats they got.
He says "Hard had 1, just like me." So for me that sounds like it wasn't cheating.
Yet, you failed to consider what I said.
Just because the AI is on par with you, doesn't mean it isn't cheating. AI's suck by comparison to a human mind and cheating is the only way this game has for the AI to make life miserable for you
Carbon's experiment is also extremely limited and highly flawed and should not be used as the ultimate example, merely proof of concept. He likely didn't make all of the AI's be the same race, probably left the AI's at random rather then setting the AI's to be all the same personality or more appropriately all to researchers like he was going to play and only gave enough time to compare the ability for himself to get 1 single research out, failing to realize you can do multiple (up to 5) researches at the same time and he's only comparing the values for combat research and not taking into account any of the research done in any of the other 4 trees, namely defense, non combat, fleet, and diplomacy (which isn't even shown on the end game summary). He also fails to mention any research done by the unfair level AI, which is the biggest clue to me that he failed to eliminate as many variables as possible and include as many research trees as possible to get the most accurate comparison.
Unlike Carbon, I set up a similar research experiment where there was myself, and 5 AI's. One each of normal, hard, unfair, cruel, and vicious. All 5 were set to TEC so they'd all have the same research costs and times, and set to be researchers so that research would be their priority as it was my priority in this test, these were the results I got, not including any of the researches done on the diplomacy tree since those aren't shown by the end game summary and therefor, impossible for me to know how much the AI's actually accomplished there.
After 6 minutes with all game settings set to normal, I was able to not only build one of each research station, but also complete 1 basic level research item from all 5 trees simultaneously. I let the game run a couple minutes longer just so that I'd know the 6 minute mark would have all of the values recorded and listed in the end game summary.
Combat Tree:
Me: 1
Normal: 0
Hard: 2
Unfair: 3
Cruel: 13
Vicious: 21
Defense Tree:
Me: 1
Normal: 0
Hard: 1
Unfair: 2
Cruel: 3
Vicious: 4
Non Combat Tree: (If I were to only compare the results from this tree like how carbon did with his combat tree example, then there's no way I could prove the AI was cheating with research since Cruel had less then me whereas the others had more.)
Me: 1
Normal: 2
Hard: 2
Unfair: 2
Cruel: 0
Vicious: 5
Fleet Tree: (Same with this tree, we all had the same.)
Me: 1
Normal: 1
Hard: 1
Unfair: 1
Cruel: 1
Vicious: 1
Diplomacy Tree:
Me: 1
Normal: ?
Hard: ?
Unfair: ?
Cruel: 0 (I know this because it had no non combat research stations built)
Vicious: ?
Overall research completed, not including diplomacy research:
Me: 4
Normal: 3
Hard: 6
Unfair: 8
Cruel: 17
Vicious: 31
You can see the trend even without knowing what was done in the diplomacy tree, since there are only 2 possible research items with only 1 station built like all of us but cruel had, the numbers aren't going to be all that different. Also consider that all 5 AI's built ships at the same time as researching and I did not. Normal doesn't cheat, which is validated by the fact it completed less research then me, but since it built a couple of ships one must understand that the ship building robbed it of resources and prevented it from keeping on pace with my research output. Hard had more ships, unfair had even more, cruel had even more then that, and vicious had the most, as one would expect, and they also completed more research. This proves that not only does the cheating start at the hard AI level since it was impossible for me to keep up with it even though I did everything possible to get my research out as fast as possible and they did not (read that as I built my research labs right next to my builders so they had almost no travel time and queued up my research so it would start the instant the labs were finished, whereas the AI's built their stations at random places adding travel time), it also confirms that they get research speed and ship building time bonuses, or cheats, if you prefer to stick with that term, since the harder the AI the more it pumped out, as well as the resource bonus to pay for all of that stuff.
And I can't compare after a a longer time for maybe the AI has more gold income for he has focused on economy while I have focused on something else.
Not true at all. This actually goes back to an earlier comment you made because I know you're looking at the wrong stats....
Like I said the gold income seems to be the same as mine.
Credit (not gold, btw) income is a horrible stat to look at. All of the income rate stats - credits, metal, and crystal are going to be the exact same in a game where no expanding has had a chance to occur and vastly different in the later stages because it's affected by far too many things all at the same time. In my 6 minute research experiment, all 6 of us had the exact same credit income rate. HOWEVER! Planet income was vastly different. Normal and myself were only off by about 100 credits from each other, both around 6000, which one would expect since we're playing with the same values. Vicious on the other hand had gained over 36000 credits (six times as much as us non cheaters) and considering it had only done 5 non combat research items, I doubt that 36000 came from maxing out it's economy on the only planet it held. The other AI's fell appropriately in between us right where one would expect them to be at roughly 9000 (one and a half times as much) for Hard, 12000 (twice as much) for Unfair, and 24000 (four times as much) for Cruel. Metal and crystal were similar.
When looking at the summary screen you can compare yourself to the cheating AI's by comparing three groups of things:
Resources gained from asteroids/credits gained from planets vs spending.
The 4 of the 5 research trees you can look at.
Used ship slots.
Why those three things? The first one let's you know just how much the AI is cheating in it's economy. It will outpace you the entire game until you crush it's economy by taking away it's planets. If your economy matches or exceeds a cheating AI's spending, then you've got no problems. The second one let's you know how much it's cheating in getting it's research done. It will get most of it's research done before you, before it decides to slow way down. All that matters is how far ahead of you it is when it sends it's fleet after you. If it's a lot, you're in trouble. The third let's you know just how big the AI's fleet is and how quickly it's replacing it's destroyed ships. If you can keep up, then you're going to be just fine.
Now more specifically to your comment, if the AI researches non combat heavily (expanding it's economy) and you research fleet and combat systems to make better, longer lasting, harder hitting ships, you still lose because it just spams ships at you and wins a war of attrition until you get over there and do something about it's economy... You can't honestly sit there and tell me you can't figure out how much the AI is cheating when you compare the correct stats. And you don't have to use solely the three I mentioned. There's also cap ships, planet's owned, and structures built. Those three are good to look at to see how much it's cheating too. 4 planets but a better econ then your 12 is a pretty good clue. 14 structures added immediately following adding another planet let's you know it's got money to burn. Losing 4 caps and replacing them right away let's you know it's mad and going to come back real soon. Well, not really, but the AI doesn't just let it's newly build caps sit around forever.
If you really want to see just how pathetic the AI is even at vicious, load up the dev.exe, start a game, press control+shift+> at the same time, and then press 'a' twice. You'll have all of your research completed immediately and you'll be given an absurd amount of resources, all of those pesky things like logistics and tactical slots Capital ship crew maxes and fleet supply maxes are a thing of the past, everything instabuilds, and best of all, rapid fire novaliths... (But maybe that's cheating just a little too much, you know.) Then you can go and stomp all over the AI because there's no way it can beat you in research or economy. All it can hope to do is beat you with ship spam, which shouldn't matter if you give yourself 10 or so frigate factories and a few cap ship factories in the same gravity well and shift click your way to nirvana... Then it's just spam vs spam until you or the AI gets bored and quits, or in my case eats about 200 warheads from a dozen Novaliths set to fire like an assault rifle. >=D
