This game is just fabulous!

[nas1m]

TigerBot - 600 / 4 RNG   - i dont care that its almost harmless. I simply run out of ammo if there are too many of these. This  bastard doesnt belong to "smaller".

So maybe drop HP to 400 and move it to Middling?

DoomBot - 800 HP / 7 RNG   - how the HELL is that monstrosity considered "Middling"?  It has more hp AND more range than the rest of its class, and it's quirk also makes it STRONGER without giving it a weakness.

Yea, that should move up to Larger, and maybe come down to 500 HP?

ThiefBot - 40 HP / 5 RNG   -  5 range is not too much, ammo theft is not that terrible, especially on harder difficulties where everything one-shots you. Why then it has so little health?

So possibly buff the hp to dumbot levels and move it to Smaller?

AmmoBot - 60 HP / 3 RNG  - even at 10x, I'm not sure this guy's virus cost is high enough. Unlimited plasma cannon ammo make almost any mission trivial. No other bot offers an advantage as huge as this one.

Yea, I think it should probably be changed such that if it gifts ammo to a player exo it only increases each weapon by 1 unit of ammo and then the AmmoBot selfdestructs.  Possibly bring the virus cost back down some, dunno.  I'd like virusing these things to be an interesting choice (so not making them immune to virus) but I don't want them to provide an "asymptotic" benefit, for lack of a better term.

WyvernBot - 800 HP / 7 RNG - I know even the description says its almost a boss-tier bot, but i really dont like its design. Almost all other bots have quirks, and often weaknesses that can be exploited. This guy is just all round super-strong, with no weaknesses or any interesting mechanics. I would be ok if it was named "PeonBot, or SoldierBot" and was a 200 HP / 5 RNG boring bot with no powers. Why make an all around boring unit and then give it the stats of (almost) a boss?

I'll need to review that one and figure out if there's some quirk that's just not working or whatever.  Either way the HP could come down to 600 or 500 to take a bit of the "bullet sponge" edge off. 

I can imagine how hilariously easy the final battle would be if the "army" was made of the bots with lowest HP: DumBot + Invigorated Dumbot + NinjaBot + AmmoBot + SilenceBot. I am absolutely certain i could beat that without breaking a sweat on expert, possibly even on misery.

Suggestions on changing the end-game-bot picking logic are welcome.  Possibly aiming for a certain target HP per 100 bots based on an expected ratio of smaller:middling:larger?  Not that it'd follow that super-strictly to avoid hamstringing the randomness, but also not letting it go to the extreme outliers.

The suggested bot and bot selection changes look good to me. Definitely an improvement, I think.
The same goes for the WyvernBot needing a bug that can be exploited. It needs one. Badly .

[nas1m]

Just a quick follow up after playing some more (on Expert)

TigerBot - 600 / 4 RNG   - i dont care that its almost harmless. I simply run out of ammo if there are too many of these. This  bastard doesnt belong to "smaller".

So maybe drop HP to 400 and move it to Middling?

Yes. Definitely. Please. If my current game is any indicator moving it to Middling (and assuming this causes fewer of them to be spawned) is even more important than lowering their individual shields. Even though dropping its shields slightly is likely  good idea as well .

For reference:
I am in my third mission in my current Expert game and TigerBots are at level 3, which means 6000 shields, while the best damage I can do is around 1100 using my Siege Exo. I have currently 8 shots that are able to deal this much damage. The grouping in Smaller caused a significant amount of TigerBots (10+) to be spawned in my current ScienceStation map - with way more shields than are fun (or even possible) to wear down.   

[nas1m]

TreadBot could definitely stand a change as well. On level 4 these sport 2202 shields on Expert while being part of the "Smaller" category, which makes fighting them a slugfest (and ammo sponge). I think they are fine if moved to "Middling", though (presuming these get seeded significantly less often).

As an afterthought: Maybe moving all bots sporting a base shield value of greater 100 to Middling as a rule is in order? Presuming your "Smaller" category matches the proposed "Chaff" category, that is. If not, adding one for DumBots & Co might be an alternative. Multiple thousand shields just are not very "chaffy" in my opinion...

BTW, what are the relative pick rates for the various categories? Just to get a mental image .   

[keith.lamothe]

BTW, what are the relative pick rates for the various categories? Just to get a mental image .   

The RelativeFrequency column on the table I posted earlier is basically your answer   The values range from 11 for a relatively-rare bot to 49 for a very-common one.  Those frequencies equate to the bot's size class, and the Smaller/Middling/Larger categorization I gave earlier is basically an aggregation of size classes.

Anyway, as that table shows, the TreadBot is one of the two bots with the very highest frequency (49), which doesn't seem appropriate at all   Given the health, at least.

[nas1m]

ShieldBot might be worth a look as well. On level 3 these got 10000 shields in my current Expert game on roughly day 5 or 6. That's - quite a lot. Maybe base shields of 1000 (as indicated by the wiki) is a bit steep? Don't get me wrong these should be hard to kill - but maybe not that hard that early in the game...

The RelativeFrequency column on the table I posted earlier is basically your answer   The values range from 11 for a relatively-rare bot to 49 for a very-common one.  Those frequencies equate to the bot's size class, and the Smaller/Middling/Larger categorization I gave earlier is basically an aggregation of size classes.

Anyway, as that table shows, the TreadBot is one of the two bots with the very highest frequency (49), which doesn't seem appropriate at all   Given the health, at least.

Ah okay, I see. These look roughly sane to me - aside from the already mentioned TigerBots and TreadBots , which are much too common for the amount of shields they have.

Another one that stands out to me are DoomBots. Are these really supposed to be more common than e.g. BlunderBots and GaffeBots (35 compared to 23)?

[keith.lamothe]

Are these really supposed to be more common than e.g. BlunderBots and GaffeBots (35 compared to 23)?

"Supposed to" is a complex concept   I have no idea what the intent was (I didn't specify those stats, and probably didn't implement them either), or if the actual implementation followed the intent.  At this stage the original intent is of secondary relevance; what's more important is what appears to be a good next step for the game.

Which seems to include making DoomBots not more common than BlunderBots on levels with both of them

[keith.lamothe]

On the WyvernBot, I checked and it does have one special-behavior flag: GetsFreeShotAtFlavorObject (basically this means destructible cover objects).  This is similar to the ThunderBot except it doesn't make those cover objects always explode violently.

If that's not special enough, perhaps it can cause the destroyed cover object to be replaced by a random really-weak (i.e. DumBot) enemy?

[nas1m]

On the WyvernBot, I checked and it does have one special-behavior flag: GetsFreeShotAtFlavorObject (basically this means destructible cover objects).  This is similar to the ThunderBot except it doesn't make those cover objects always explode violently.

If that's not special enough, perhaps it can cause the destroyed cover object to be replaced by a random really-weak (i.e. DumBot) enemy?

I think the special ability is fine, per se. The problem I see is that the bot could use some additional quirk that can be exploited by the player to compensate for it's (almost) boss-like stats. As for what is fitting here I am at a loss at the moment .

Alternatively you could dump him down to to, say, half his stats in terms of shields but have him destroy multiple cover objects? He would still be dangerous for its damage and cover clearing ability, but not as presumably OP anymore...