We’re past our next tournament, so the GDT and Balance team have chosen to touch up the game in a few aspects, primarily concerning T1 bots.

Faster pawns, slower grunts

The Grunt‘s speed was reduced by 3, and the Pawn‘s speed was increased by 3 as well. They wre previously equal at 84 speed.

Note - I don’t know the full reasoning behind this change. What follows is my opinion and attempt to understand it.

The Grunt has 230 range, compared to 180 for the Pawn. This means that, as long as the Cortex player reacted within the 0.6 seconds necessary for the Pawn to close the distance into firing range (assuming a stationary Grunt and neglecting acceleration), the Grunt could infinitely kite its more expensive counterpart and thus hard-counter it.

If we now assume the Grunt kiting backwards right from the get go, the Pawn will cross the 50 units of distance using its 6 speed advantage within 8.3 seconds, and it will then take take down the 270 HP grunt via its 90 DPS in over 3 seconds (due to imperfect accuracy, kiting advantage, etc).

Overall, this should not diametrically change this particular interaction. It will definitely give more of a reason to actually make Pawns from time to time, and the Grunt should still remain a strong option. I get some of the panic in #main chat, but I think it’s good to try it out first and then give feedback.

Demon build cost (time to build) adjustment

Asked about the Demon’s cheap build cost of 90 000 (time to build = build cost / applied build power, assuming sufficient resources), prolific developer Hornet replied “unsure, it was 150 000 at one point“. One of the many wondrous accidents that inevitably happen during development, it was changed and forgotten about.

The build cost was thus increased to the midway point of 120 000, a 33% slowdown to Demon summoning efficiency. You will need more grimoires construction turrets to spam them out now.

Plasma bot social distancing

You’ll find that your Maces and Thugs don’t clump up as much, and this may make them actually a fair bit better.

Everyone knows that Thugs are very strong and Maces are pretty bad, right? They should be the “same” unit, but the former just feels much better to play with, even if neither is very strong in the current Rocket bot dominated meta.

The common folk knowledge answer on the cause of that difference is their build cost and HP. Maces have 900 HP for a build cost of 2200, which comes to 0.45 health repaired per second per unit of build power. Thugs, on the other hand, have 1000 HP for a build cost of 1970, giving us 0.508 for that same quantity - a 12% increase. Not massive, but decently larger.

But there’s a larger difference that takes a lot of deep digging to uncover, and while figuring it out, investigative reporter Hornet noticed something interesting about the situation. Consider this:

If you have a very tight ball of either Plasma Bot, because of how their models work, many of the actual cannons-as-dictated-by-model will up end inside the collision volumes (purple) of another bot. And there they hit an engine bug that prematurely ends the search for viable targets, because they do not attempt to rotate towards viable targets (either by shifting their turrets around, or by arcing the cannons upward).

Zecrus noticed that this is further exacerbated by this difference between the two models, so the Mace is hit harder by this:

Here’s an example of the change, with changed Thugs at the top and unchanged Maces at the bottom, in equal numbers.

There is an engine pull request that would improve this situation, but for now, both Thugs and Maces were tentatively shifted to a larger movedef to prevent this.

In summary: your Plasma bot per meter elmo unit area count will decrease, but the DPS might actually increase. This may actually make plasma bots more viable. Try it out!