Can't consistently win hamlets against AI

Share and discuss strategies for playing the game, and get help and tips from other players.

Moderator: Forum Moderators

Post Reply
User avatar
Oflameo
Posts: 21
Joined: January 31st, 2012, 12:17 am

Can't consistently win hamlets against AI

Post by Oflameo »

I am playing against the Experimental AI on the Hamlets with me as Drakes and the AI as Northerners. I gave the AI a +9 gold handicap, but I can't beat them consistently. I have two winning and losing games uploaded so you can evaluate where I am sloppy.
Attachments
2p_—_Hamlets_Fail_0.gz
(32.9 KiB) Downloaded 524 times
2p_—_Hamlets_win_1.gz
(28.22 KiB) Downloaded 559 times
2p_—_Hamlets_win_0.gz
(25.36 KiB) Downloaded 546 times
2p_—_Hamlets_Fail_1.gz
(37.41 KiB) Downloaded 546 times
Last edited by Oflameo on September 28th, 2016, 5:26 pm, edited 2 times in total.
Velensk
Multiplayer Contributor
Posts: 4002
Joined: January 24th, 2007, 12:56 am

Re: Can't consistently win hamlets against AI

Post by Velensk »

I can watch these replays another time but I'd like to make a couple notes here before I even do that.

A +9 gold per turn handicap is significant, especially in that match-up. The AI isn't anywhere near as good as a human but even a mediocre player with that kinda handicap would have pretty good odds vs an expert player.

Playing against the AI is different than playing against a human and the skills required to win are different. If you're looking for the kind of feedback that will help your play against other humans I might be able to drag up a few mechanical/tactical/calculation comments but almost nothing I tell you will be useful on a strategic level under normal conditions.

Basically this style of asking for feedback will yield only advice on the uttermost basics of play. I'll still do it (either tomorrow or the day after) but understand what you're getting.
"There are two kinds of old men in the world. The kind who didn't go to war and who say that they should have lived fast died young and left a handsome corpse and the old men who did go to war and who say that there is no such thing as a handsome corpse."
User avatar
Oflameo
Posts: 21
Joined: January 31st, 2012, 12:17 am

Re: Can't consistently win hamlets against AI

Post by Oflameo »

I noticed that I didn't upload Fail 0 yesterday and I uploaded win 0 twice instead. I just fixed it.
User avatar
Eagle_11
Posts: 759
Joined: November 20th, 2013, 12:20 pm

Re: Can't consistently win hamlets against AI

Post by Eagle_11 »

Just rotate the frontline, AI can only suicide into your defenses.
Velensk
Multiplayer Contributor
Posts: 4002
Joined: January 24th, 2007, 12:56 am

Re: Can't consistently win hamlets against AI

Post by Velensk »

On Win 0

Turn 1: Your recruit is poor for the situation. It's important to grab villages quickly but you don't need that many scouting units for fast village grabbing and the scouting units provide less raw power per cost. Your early challenge will be overcoming the first night attack and for that durability and killing power are what you want.'

Turn 4: Your moves on this turn would be a mistake against a human opponent but aren't too bad an option against the AI as it won't exploit a few of your decisions to lock your leader and your economy down and take a very favorable nighttime engagement (and also will let that saurian get past.

Turn 5: Considering how well that ended up working out for you, this game should pretty much over.

Turn 9: Yeah, that's pretty much it.

On Win 1:

Turn 1: Again, recruit raw power (while still having an optimized village grabbing plan) for besting the AI.

Turn 3: Stealing the AI village would have been a mistake vs a human player. The extra income you get wouldn't have been worth the loss of a glider and xp for the enemy. A human would have used ZoC to prevent your retreat and killed the glider on the next turn.

Turn 4: Against a human, going mass saurian vs orcs would be a poor choice. Saurians can do ok so long as there is terrain for them to hide in but the more you have, the more inevitable it becomes that they'll have to stand in the open and orcs murder saurians in the open. Vs the AI it doesn't matter and is in fact profitable for making your counter attack fearsome.

Turn 5: Temporarily grabbing villages generally isn't worth sacrificing units unless you are trying to keep his troops busy clearing out your village stealers as opposed to fighting on some other front. It's not worth trading a saurian to control a village for 3 player turns (+3 on your turn, -3 for him on his turn, +3 for you on your turn).

Turn 6: This game is pretty much over at this point.

On Fail 0:

Turn 1: Same note with the recruiting.

Turn 5: Your tactics here have put a lot of saurians into the open at night. Fortunately the enemy doesn't have a strong grunt followup like they would if it were a human but even with the few units that can reach them, that's going to be painful.

Turn 6: Your tactics are killing you. Not only are you leaving yourself horribly exposed to a counter attack (fortunately, your opponent got unlucky there) but you choose to keep your leader in the fray. It's far more important that you get some reinforcements a turn sooner than that you kill that archer that turn.

Turn 7: Ow. That auger missing cost you a lot. Of course, it's better if you don't have to put so much on one engagement but that was about the best you could do given the set-up.

Turn 10: You're simply too far behind at this point. Drakes fighting northerners really need to avoid the situation where they're constantly having to retake their villages.

On Fail 1:

Turn 1: Same note with the recruiting.

Turn 4: Your actions on this turn demonstrate a poor understanding of the situation. You easily kill the grunt on the left flank (and rightfully so) but then you start sending troops running across the field to put on pressure on the left. As mentioned before it's not worth losing troops to hold a village for a very short period of time so by doing this you should be committing to an actual offense. The timing for this offense is incredibly poorly timed though. It is about to be night and while your saurians gain some bonus, the entire orc army is stronger at night and your drakes lack killing power. You are being pressured hard on the right flank and at night that fight is one that is far too fair, possibly favoring your enemy slightly. You've just sent your leader out so reinforcements are a few turns away and your leader won't be able to help you defend and in the meanwhile your opponent will be able to force an engagement (at least on the glider) that you can't escape from while constantly providing reinforcements to the conflict you've started at his village. Your odds on every front aren't great and if you lose any of them your enemy has better followup than you do.

Turn 5: I don't know if you know how lucky you got this turn both on the strikes and the enemies choice of action.

Turn 8 That glider up top has certainly paid for itself however it only has because the AI has been playing very poorly. In the meanwhile your situation down south is kind of grim. Even in a normal match without a +9 income bonus, the situation that the orcs need to avoid above all else is the one where they have to spend all of their efforts running around chasing orcs off their villages. The snowballing effect against the limited number of units creates a painful dynamic where every night costs just a little bit too much and the drakes never quite recapture momentum. Even against the AI I don't see this changing.

Turn 14: At this point, it probably feels to you like you are stabilizing. This is to some extent true but you don't really have the balance to attack yet. You should focus on chomping down on the scattered attackers that are streaming in, you'll find your opponent soon would have no options. By attacking him and allowing the forces on the right to roam free you give him some options.

Turn 15: Again, it's not worth holding a village for an extra turn if it means losing a unit. Also, putting the glider there gives him a very easy shot at leveling a archer to a crossbow which is something you really don't want to have to deal with (of course, the AI is too dumb to take advantage of this).

Turn 17: This is pretty much over. That force you have on the left is strong but you just aren't going to be able to fight an unending tide of orcs and your leader has no way of getting all those villages back and saving himself.

OVERALL: Honestly, every part of your play could use significant improvement. I would start by focusing on your strategic sense of the game and priorities.
--It's good that you understand the value of villages however for the moment I'd suggest focusing on preserving your units, especially as while playing drakes. Units are just as much a resource as gold is but if you play better than your opponent the value of the units becomes much greater than their gold cost suggests. A stronger tactician in Wesnoth can with equal cost of units trade units incredibly cost efficiently.
--In any match-up where one side is one alignment and their opponent is the opposite alignment, the side that can force a major engagement during their time of day will almost certainly derive such an advantage that their opponent will almost certainly lose. Focus on trying to force your opponent to fight at a time of day favorable to you.

If you want to be able to consistently beat the AI at this match-up, I would suggest you take the following strategy: Recruit mostly drakes and grab your villages early. Your enemy will attack you as the first night is falling. Judge how well you can hold them off without losing any units. BE WILLING TO RETREAT, even if it means losing villages and just stall until the second day without worrying about taking offensive actions. As the second day is starting launch a counter attack and don't stop until you've destroyed all enemy units on your half of the map. At this point, if you have enough strength to continue immediately press to the enemy side but if not fall back heal up and repeat. Eventually you will have an army worth mentioning and your opponent won't.
--It's not quite this simple against a human but you can rely on the AI to continue to march it's forces relentlessly towards you even if it's a bad idea. The AI is bad at retreating thus you can afford to simply let it have a couple villages for a few turns so long as you can get a massively favorable engagement on that second day with your leader and a ton of well managed drakes.

EDIT: A note on that strategy, it can be worthwhile to send a scout, not necessarily to grab villages but just to lure enemies that would be attacking you that first night away. However, it should be possible to do this without sacrificing the unit by placing it on a village where it can be flanked.
"There are two kinds of old men in the world. The kind who didn't go to war and who say that they should have lived fast died young and left a handsome corpse and the old men who did go to war and who say that there is no such thing as a handsome corpse."
User avatar
Oflameo
Posts: 21
Joined: January 31st, 2012, 12:17 am

Re: Can't consistently win hamlets against AI

Post by Oflameo »

I could have to do with my play style or the Experimental AI's complete lack of fear, but I can't give up villages early in the game and expect to put up a good enough offense.

Pretty much I can't win a theater unless I have at least an equal number of units in the area. If I retreat too early in the game there will be too many Northerners than I can kill and and I won't have money to replace units to at least half of the rate as the Northerners. The AI doesn't even recruit goblins so it doesn't make a Clasher even worth recruiting IMO. If I have a unit or two on the other side of the map, it will create room to retreat because all of the Northerner units won't be piled on top of me so I can't still dominate them unit to unit I have two or three units leveled up.
Attachments
2p_—_Hamlets_win_2.gz
(21.72 KiB) Downloaded 581 times
2p_—_Hamlets_Fail_4.gz
(31.54 KiB) Downloaded 549 times
2p_—_Hamlets_Fail_3.gz
(18.4 KiB) Downloaded 565 times
2p_—_Hamlets_Fail_2.gz
(30.37 KiB) Downloaded 509 times
Velensk
Multiplayer Contributor
Posts: 4002
Joined: January 24th, 2007, 12:56 am

Re: Can't consistently win hamlets against AI

Post by Velensk »

For what it's worth, I actually tried the challenge you are doing and I found that I couldn't do it consistently either. A couple notes on this:

-I don't know if you've seen this in your play, but for me I won every game I was first player but I lost quite a few of the ones where I played player 2. Effectively, the enemy attack comes a ToD segment and turn later when you're playing as player 1 and this makes all the difference. I think that getting more saurians for a nighttime counterattack as player 2 might have some merit but this is still is a poor map/matchup for it.
-The experimental AI is a full handstand above the default AI in strength. I was impressed not only with it's better movement/recruit patterns but by the fact that it understood the value of grabbing villages as orcs and was actually capable of making a few sensible retreats. It's still no match for a human player but the +9 income per turn puts it into the range of a human.
-When I won it felt easy, when I lost it felt impossible, I think it's because the nature of an enemy with such an income advantage that makes momentum paramount.

I don't know if this helps but here's a replay of one of my wins. I recall feeling like I'd gotten lucky this game but I expect I would have won anyway. As I stated earlier, I won all the attempts where I was player 1.
Attachments
Example.gz
(24.34 KiB) Downloaded 513 times
"There are two kinds of old men in the world. The kind who didn't go to war and who say that they should have lived fast died young and left a handsome corpse and the old men who did go to war and who say that there is no such thing as a handsome corpse."
Post Reply