Co dev-lobsters wanted !

I might be interested if you can explain the methodology.

Like I'm not even sure what these numbers even mean:

Can you explain what you mean?
I'm pretty sure 5 Glaives will win against a single Blitz for example.
I don't think Blitz is a good counter to Glaives.

Quite a few matchups don't reduce neatly to a single number. They depend strongly on...

Terrain: One Mace reliably beats five Ducks on flat land, although it's not a hard counter. Two Ducks hard-counter a Mace in deep water, killing it quickly without taking any damage in return.

Scaling: Three Glaives can beat a Venom by attacking from an arc. Against ten Venoms, even a hundred Glaives do poorly, although the hundred Glaives can still overrun the Venom ball at great cost.

Combined arms: Ravagers beat Fencers handily. Placeholders can only inconvenience Ravagers. But the combination of a few Placeholders and a ton of Fencers does considerably better at killing Ravagers than either pure composition.

Micro degree: Ten Fencers being actively and skillfully microed beat a pair of similarly skillfully microed Ogres without taking a shot. Ten Fencers on attack move or sitting still only barely beat a pair of Ogres on attack move.

So what are your assumptions?

What does 0 or 2 mean?
I mean, units without weapons, like builders, should arguably have 0 against units that can kill them.
Or do they get a higher value if they are extra tanky or faster than the attacker?

quote: Specifically regarding the glaive vs blitz interaction i think large number of blitz say 5 vs 20 glaives will win the engagement, no ?

Blitz has no AoE(Area of Effect/splash damage), but how Blitz scales against Glaives is still quite complicated. Each lighting shot from the Blitz only hits 1 target.

It just takes 3 Glaives to defeat a single Blitz if the Glaives are in range of the Blitz.
However, if the Blitz can 1v1 the Glaives, it takes something like 8 Glaives to kill a blitz.
Yet again, since the Blitz has longer range than the Glaive, it can also be able to kill the Glaive before it enters range.
Glaives are faster than Blitzes but if the Blitzes can kill the Glaives before they can fire back, Blitzes can kill infinite amount of Glaives with no losses.

I'm not sure which scenario is the most realistic or important when you want to decide the match-up between Glaive and Blitz in an actual game between players.

Kodachi has a more favorable match-up against Glaives than Blitz does.

If 3 Glaives can get in range of a single Kodachi and they are all spread out enough to avoid the splash damage, they will get a costeffective trade, only losing 1 Glaive in the process.

However, while it might be feasible to chase a Blitz and get in range and kill the Blitz costeffectively, chasing a Kodachi with Glaives is basically suicide and it is basically impossible to make a costeffective trade against a retreating Kodachi unless you can surround it first.

Now, if the objective is just to protect a target against a single kodachi, 3 Glaives can do that, because the Kodachi will lose if it just drives straight into the Glaives.

I'm not sure if all those scenarios can be rolled into a single number, a single ratio, a single dimension.

quote: The scoring really means: If they have X and I build Y, will Y deal with X?

Yeah, but that doesn't have a singular answer. It depends on other circumstances and the stage of the game.

If you use Glaives for defense against lone or small numbers of Kodachi, Glaives are good enough.

If you use Glaives against a large amount of Blitzes, they will simply die 1 by 1 before they can get into range.

I'd suggest that the full 2.0 advantage be reserved for matchups where the advantaged unit can both force a fight and get reliable one-sided attrition despite a large resource difference. Harpy vs. Reaver would be a good example, as a ball of 4+ Harpies can chase a large group of Reavers and pick them off one at a time, and the Reavers only get to shoot back if the Harpy player makes an error.

I'd suggest that matchups where the disfavored side can elude the favored side indefinitely be capped at about 1.5-0.5. In order to elude properly, the disfavored unit must be faster, and have vision range exceeding the favored unit's attack range by enough that the favored unit cannot close from vision range to attack range in the time it takes the disfavored unit to turn around. This seems particularly important for situations with AA units vs. a few medium-to-heavy ground units with lots of air support; AA that can't be chased down gets a comparative advantage.

Finally, I'd suggest that the mere potential existence of favorable terrain somewhere on the map that can radically improve a matchup should be worth 0.1, if you can't check directly for relevant terrain.