Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Iduno]

Thank you for the numbers. I thought they had set the AC/5 to 9 (which would be quite OP), but it seems I've misremembered!

With something like an Awesome, I think I'd pass on the AC/10s and run 3 AC/5s and a PPC. 34 damage for even more heat neutrality than before. It comes with a bunch of little tradeoffs (more spread damage, more chances to hit/miss, ammo explosions, more firepower when engine crit.)

And more TACs.

[Rince Wind]

There is an option to turn off TACs, which is always enabled in my games.

Re: Princess is slow: I am seriously thinking about upping the quality of the enemy troops in my current mission, because the Blakists send mostly hovers now. And those take princess ages to move, because of their speed.
If I just change the ratio of mech and vee lances I'll probably just end up with a bunch of 6/9/6 and faster mechs that take just as long to process.
But hey, look at my two new only slightly used Regulator IIs! 10/15 Gauss Rifles on hover skirts. They did cost my a Duan Gong I never cared about and a PXH 1c (special). I love that thing to bits, even if it is not a production variant it is one of my favourite mechs. But I forgot that that omni Blackjack could just flip its arms...so yeah, it is not only hovers.

[Robsoie]

I really hope they'll find a way to have Princess processing its decision and moves much faster, on some quick and simple battles the bot sometime can play a turn fast but on some turns it's very long (and all on small maps and only 4vs4 mechs).

I don't want to imagine trying to play large scale fights on large maps.

Anyways, my favorite thing in Megamek is robot karate like on the scenario with kurita vs davion

Spoiler (click to show/hide)

[Teneb]

I decided to give this a spin... and had the AI handily beat me without losing a single 'mech.

I did notice the long decision times mentioned above though.

[Iduno]

I decided to give this a spin... and had the AI handily beat me without losing a single 'mech.

I did notice the long decision times mentioned above though.

How well balanced was the fight? If you start with some of the training scenarios (some are trash), you should be able to handle Princess in a fair fight, and from there be able to consistently win even fights. The AI is good enough that you can't really beat it at much more than even fights unless you understand the decisions it's making.

[Xardalas]

Oh man, it always bring joy to my heart to see this thread on the front page.

I really hope they'll find a way to have Princess processing its decision and moves much faster, on some quick and simple battles the bot sometime can play a turn fast but on some turns it's very long (and all on small maps and only 4vs4 mechs).

I don't want to imagine trying to play large scale fights on large maps.

Anyways, my favorite thing in Megamek is robot karate like on the scenario with kurita vs davion

Spoiler (click to show/hide)

I had a large scale battle once with my mercs that took me about four hours to play out. Most of that was waiting on the AI. Dear god base attacks were painful. So many units.

[MrRoboto75]

I gave the first training mission a shot.  Wasp+Griffin vs Stinger+Wolverine.  Oddly, the griff and wolverine are the ones in the new BT starter box.  Should be said its likely my megamek version is probably out of date.

Anyways the first half of the battle was rough.  Prin consistently got initiative, and the griff you get only has long range guns, Prin prioritizes just sitting inside the griff's minimum range.  The wasp ended up dying without doing all too much, but in doing so the griff could get somewhere where it could actually try to hit something.  A couple lucky hits dets the stingers MG ammo, which is naturally the most destructive force in all of battletech.

From there on after a while the wolv kinda gave up.  It'd try to trade blows at med/long range instead of close in.  9 vs 7 to hit meant it was losing.  It tried to close again but I just jump between forest blobs and over a lake.  Eventually it parked itself away from me with its back facing me, and it wouldn't bother shooting me.  It died eventually.

I think the dead stinger spitting out a mechwarrior confused the bot.  Its an infantry unit that doesn't really do anything.  The wolverine scooped him up for whatever reason, and I guess had no idea what to do with it.

[Rince Wind]

I don't think the AI picks up MWs intentionally.

At least it doesn't shoot them anymore!

You can use them as initiative sinks (if they are yours) or pick them up yourself. In extreme weather that is important. Not that the AI knows, it just dumped a squad of infantry to die in the freezing cold. Poor bastards.

I wish there was a "white flag" option for vees, where they give up and aren't targeted anymore, but stay within their ride.

[Iduno]

I gave the first training mission a shot.  Wasp+Griffin vs Stinger+Wolverine.  Oddly, the griff and wolverine are the ones in the new BT starter box.  Should be said its likely my megamek version is probably out of date.

Anyways the first half of the battle was rough.  Prin consistently got initiative, and the griff you get only has long range guns, Prin prioritizes just sitting inside the griff's minimum range.  The wasp ended up dying without doing all too much, but in doing so the griff could get somewhere where it could actually try to hit something.  A couple lucky hits dets the stingers MG ammo, which is naturally the most destructive force in all of battletech.

From there on after a while the wolv kinda gave up.  It'd try to trade blows at med/long range instead of close in.  9 vs 7 to hit meant it was losing.  It tried to close again but I just jump between forest blobs and over a lake.  Eventually it parked itself away from me with its back facing me, and it wouldn't bother shooting me.  It died eventually.

I think the dead stinger spitting out a mechwarrior confused the bot.  Its an infantry unit that doesn't really do anything.  The wolverine scooped him up for whatever reason, and I guess had no idea what to do with it.

Light mechs (stinger and wasp) die fast if they get hit, but should also be able to consistently get a +3 or 4 to hit them by running around a lot and finding woods. The griffin is sort of a long distance mech, trying to stay just outside of what the enemy can hit and slowly chewing away at them, occasionally jumping over a hill to avoid getting shot too much. The wolverine is a nice close-range mech. Try to stay on your side with the griffin and wasp to make the enemy come to you, or rush at them with the other set. See if that helps.

For those times when you want to hit (the other option is moving for positioning), don't jump if you can get there running, and don't run if you can get there walking. Less heat, and your to-hit modifier is lower. Light mechs usually run, because it gets them another +1 to hit vs walking, but can get the enemy another +2 to hit. Not moving at all is reasonable if the enemy can't get close enough to hit you (like when you have an Awesome).


Also, there is an option somewhere to make infantry move last, so they're less effective at initiative padding. It speeds the game up slightly.

[Rince Wind]

Not using individual initiative is making it quicker as well, but then you can't really spot for indirect fire, and might have to move vulnerable units first. Same for princess of course.

[Robsoie]

I tested with using the unofficial rules all the simultaneous move/shot/whatever and it seemed to speed up the Princess bot by a lot on the small map 3 vs 4 mechs i setup (4000BP , so i could field 2 Awesome and 1 hoover big tank vs Princess random 4 mech between 40 and 60tons) .

The map was actually horrible, tons of height 8 and 9 "pillars" all over the map
I was super surprised on how my Awesome were dealing incredibly weak damage (when they were hitting i mean as bad luck hit roll seems too common for me) , and got destroyed by the opposition easily.

At least the horrible map "pillars" played well when keeping my tank alive and force 1 slightly hurt enemy to run away from the map (why did he ever do that ?), on one of my Awesome dying turn we still managed to destroy an enemy too.

So in the end it was 2 damaged enemy mechs vs my slightly damaged hover tank, my tank hit hard and managed to make an enemy mech (a Mercury) fall down on his face , i was confident (well relatively as those damned hit rolls bad luck) i could take it out on the next turn as it was already very damaged.

I was a bit concerned about the last mech, a "hatchetman" that would probably murder my hover tank if it could catch it on melee.
So i waited for Princess to move it , then waited, 1 mn , 2 mn  ... let's browse internet ... 15mn later Princess is still on its turn to move.

Yeah the bot just bugged out completely, 2nd time in 3 battles since i have been trying this current dev version (0.47), looks like the bot need some debugging work in this version.
Oh well, at least it seems simultaneous turns really help to avoid Princess taking so damned long to move a mech on such small scale engagement.

[Rince Wind]

You can kick the bot from the game, skip the turn of the affected unit, and then reconnect it. Usually it doesn't happen on the next turn.
/who - to find out which player the bot is (should be 1 in a 1v1, but sometimes there are more bots)
/kick n (where n is the number of the bot)
/skip (or was it /skip n?)

Then there is a menu option to reconnect lost players. You'd also have to do that if you load a game.

Awesomes are awesome, but very slow, which makes them easy targets. Their PPCs don't to massive damage (30 damage if all 3 hit can be managed by some 3025 mediums), but they can reach out and touch someone pretty far away, unlike those mediums.


If an enemy is "crippled" it tries to get away (It is called forced retreat and can be turned off. In theorie you should do the same with your crippled units). The crippled state comes from different damages, if you hit it in the head with a ppc for example it would try to get away, because the head will now be destroyed if you hit it with pretty much anything but a single LRM.


Which hover did you have?
I just lost a 3058 Pegasus to consecutive motive crits. When it could only go 2/3 it got killed quickly. Because of the cold that would kill the crew within 1 turn I couldn't just abandon it either. It could have fled next turn.
I have infantry to rescue people as per against the bot rules, but 2 of the guys lost a limb, so they won't be going back on active duty anyway.

[Robsoie]

It was odd to me that this specific mech ran away as it seemed (from the health bars above them) to only be slightly wounded.

My tank a " Hover Tank (Standard) "

Thanks for those kick comands, that could come in very usefull with those bugged turns.

I have read a couple of suggestions too in order to speed up the bots :
- in the Princess settings set it at least to 70% on Berserk
- instead of using 1 Princess to control all the other mechs, use several Princess bots  to have each Princess controlling (and calculating) less mechs.

[Rince Wind]

The annoying thing about berzerk is that it'll run those LRM mechs into knife fighting range.

The different bots should help because princess first decides which mech to move, so then it has less choice.



My next fight will have a 2/3 Alacorn on the enemy side. That is three highly accurate Gauss'.
I could have cheesed and won on turn one because the weather was a tornado which means vees just die, iirc. Now after rerolling I'll just have to hope the woods will be in my favour.
Last time I faced an Alacorn there were two, and Princess hat really managed to get a good position for them, hanging back on a hill. In the valley below were, among other stuff (it was a Blake Lvl 2, so 6 units) a Demolisher, and getting 2 AC20s in the back while running up to the Alacorns would have been even worse.

[Robsoie]

Decided to play on 0.46 instead of dev 0.47 and reproduced the 3vs4 fight (excepted i generated a more normal small map), it went without bugs.

With  View -> Game Options -> Initiative Rules , having all Simultaneous (deploy/move/target/firing/physical) checked and Team Roll Initiative (in basic options) unchecked it seems Princess was much faster (i had it to 70% berserk too in case it helped), and so much more fun.

It seems they had something against my hover tank this time as i noticed 3 of the 4 Princess mechs were focusing fire on it, it unfortunately went down as i made a mistake with getting a bit out of cover to shoot at the wrong time.

The 2 Awesome survived the fight and with kicks and short range laser aimed to the mechs that were going close, and their long range missiles aimed to the mechs going to the other Awesome, it ended with a victory.

Spoiler (click to show/hide)

Poor last enemy pilot trying to escape from his downed Trebuchet