Axis AI vs Redmarkus

[Redmarkus5]

This time I will do nothing as the Soviets for the first 4 turns. Hoping to get knocked to hell by the AI this time.

[Redmarkus5]

For the avoidance of doubt...

[Apollo11]

Hi all,

ORIGINAL: redmarkus4

OK, I will try that.

Good! [:)]

I don't read everything typed here as I have a job

As most of us have as well... [;)]

If the AI "needs" 110% to attack successfully, why isn't the default setting 110%?

This is "Challenging" setting - new players should always play "Normal" (i.e. 100% vs. 100%) first to learn the ropes...

I can't imagine how many people buy the game and then give up without ever visiting the forums.

Most players play against AI and they know from manual the AI settings (if they tried "Normal" and find it to easy it is natural to try higher levels)...


BTW the forums are visited by small amount of owners of the product and even smaller amount of those do actually post - this is a known "rule of a thumb" that is acknowledged for long long long time...


Leo "Apollo11"

[Redmarkus5]

TURN 1

[Apollo11]

Hi all,

ORIGINAL: redmarkus4

This time I will do nothing as the Soviets for the first 4 turns. Hoping to get knocked to hell by the AI this time.

AI is AI - it is good but it is weaker than most of the experienced players (even with 110% help)! [;)]

Please do note that if you play the AI you are committing for long game - the AI will never rush (except for settings above 110% but those are really "murderous") - the 1941 will not be the "final decisive" year of the game against AI...

In my numerous tests (I have run close to 100 AI grand campaigns for development team) the 1942 is the best year for German AI (in it the AI seized Leningrad and Moscow the most)! [:)]


Leo "Apollo11"

[Redmarkus5]

TURN 2

Riga still in Soviet hands.

Minsk still in Soviet hands and still in supply.

AGS almost no movement this turn.

[Redmarkus5]

OK. The only points I am raising at the moment are the fact that it won't attack certain key cities (surely it can be coded to do that as part of an 'early moves' routine or script?) and that AGS in particular fails to move at all during Turn 2 and only a little in Turn 3.

If we are looking at a game where the Axis really needs to win in '41 to be sure of a win, then the early '41 AI scripts must be the most essential in the game.

A customer buying an East Front game is entitled to expect the Axis AI to blitz its way through the western USSR between June and July, no?

[Redmarkus5]

TURN 3

Riga still holds.

Minsk still holds.

Minimal movement by AGN.

Almost zero movement by AGC and AGS.

All this is without a SINGLE command to any Soviet unit since the start of the game. I just click next turn as soon as I have the screenie.

[Apollo11]

Hi all,

ORIGINAL: redmarkus4

OK. The only points I am raising at the moment are the fact that it won't attack certain key cities (surely it can be coded to do that as part of an 'early moves' routine or script?) and that AGS in particular fails to move at all during Turn 2 and only a little in Turn 3.

If we are looking at a game where the Axis really needs to win in '41 to be sure of a win, then the early '41 AI scripts must be the most essential in the game.

A customer buying an East Front game is entitled to expect the Axis AI to blitz its way through the western USSR between June and July, no?

As always in Gary Grigsby games the AI will try to follow the history as much as possible! [:)]

In UV and WitP (and new WitP-AE that was further developed without 2By3) it was the same thing...

The AI is quite capable at 110% but it will most certainly not "blitz" through Soviet defense... it is AI we are talking about after all... [;)]


Please just play at least till end of 1942 (it should not be boring)! [:)]


Leo "Apollo11"

[Redmarkus5]

TURN 4

Leo, I'm not disputing that the AI will give me a tough time. I'm just saying that it needs to be adjusted so that it behaves more aggressively (and logically) during the early war.

For example, below, this is a clear indication that the AI avoids urban areas which should actually be objectives - ports and key cities in this case. This is the situation on T4.

[randallw]

Human blitzing takes the risk of spearheads being temporarily cut off. I'm not sure one way or the other if the AI should be trying that.

[Redmarkus5]

TURN 4 - FRONT

More movement from AGN and quite a lot from AGC this turn.

AGS makes a half hearted attempt at pocketing, but fails.

Surely there is room for improvement here? This is against zero moves by myself, no reserves coming in, no new units built, nada.

Am I being unreasonable here?

[Redmarkus5]

ORIGINAL: randallw

Human blitzing takes the risk of spearheads being temporarily cut off. I'm not sure one way or the other if the AI should be trying that.

I'd be happy if it would just take Riga and Minsk...

[Apollo11]

Hi all,

ORIGINAL: redmarkus4

I'd be happy if it would just take Riga and Minsk...

AI should take both Minsk and Riga... don't you worry...


Leo "Apollo11"

[Apollo11]

Hi all,

ORIGINAL: redmarkus4

More movement from AGN and quite a lot from AGC this turn.

AGS makes a half hearted attempt at pocketing, but fails.

Surely there is room for improvement here? This is against zero moves by myself, no reserves coming in, no new units built, nada.

Am I being unreasonable here?

Yep you are... the war is not over in 1-5-10 turns (just like race is not won on 1st corner)... it is a long run... AI is programmed for long run... [;)]

Just play until at least end of 1942 and then give your observations on overall AI capatibility! [:)]


Leo "Apollo11"

[Redmarkus5]

Leo, with all due respect, if WitE is a red vs. blue sandpit game, then what you say is fine.

However, if WitE is marketed as a representation of the war between the Axis and the Soviet Union that started on 22 June 1941, then playing through to the end of 1942 in order to judge the AI's capabilities is going to be unacceptable for most consumers.

And, at the end of the day, customer expectations and customer satisfaction are the only historical facts that are going to matter

[Seminole]

Human blitzing takes the risk of spearheads being temporarily cut off. I'm not sure one way or the other if the AI should be trying that.

In my observation the best Axis players are utilizing ZOC to ensure that low morale Reds in '41 can't cut off their spearheads. The guys that get isolated are usually still learning the ropes on that aspect of game mechanics, not simply 'taking risk'.

Now later on, when the Russians can push around units trying to form the cordon, that's another ball of wax. But in early '41 when you see deep, plunging panzer offensives, the better players are creating thrusts that the Reds simply can't counter attack in a single turn.

[Apollo11]

Hi all,

ORIGINAL: redmarkus4

Leo, with all due respect, if WitE is a red vs. blue sandpit game, then what you say is fine.

However, if WitE is marketed as a representation of the war between the Axis and the Soviet Union that started on 22 June 1941, then playing through to the end of 1942 in order to judge the AI's capabilities is going to be unacceptable for most consumers.

And, at the end of the day, customer expectations and customer satisfaction are the only historical facts that are going to matter

All I am saying is that AI performance can not be judged by just playing few turns (even being few first turns at the beginning of the grand campaign) - in order to rate the AI one should play the whole grand campaign (or for the impatient until at least end of 1942 to see how things are)!


Leo "Apollo11"

[CarnageINC]

Ahhh.....forget this AI stuff Red and just pick up in January of our last game [:'(]....Come back to the darkside! 

[randallw]

In my own games the computer threatens Minsk on it's first turn and takes it on the second ( I abandon any attempt to defend it ). It would be strange if the new AI takes it later.