How do you exploit low chansey? Say I was playing someone who agreed not to use it.
I could bring:
- Back mie (preferably blizz)
- Lapras
- Probably articuno
- Probably zap because fitting good electric checks without chansey is difficult, and they often rely on mie to wall things over chansey, which makes them very exploitable to zap entering on recover
- Probably some 2018 era ref chansey team that just spam twave t1 with no fear of sing chansey.
And I'm sure better players than me have more ideas than that.
Now how do I exploit 100% chansey usage? Not a specific set, but the pokemon intrinsically. Lets assume all the sets are in their ideal proportion of usage.
Uhhhhhh...
And that's the thing. You can't really, which is why it historically had 100% usage. Dropping incredibly unexploitable pokemon doesn't make your scout less exploitable, it makes it moreso. If I replaced my Chansey 20% of the time with Pinsir, would that make me less exploitable? Obviously not.
To take this example:
This I think kind of demonstrates the problem. It's not just about how strong the move is, but how punishable it is. If they can't respond in any active way if I click blizz, whereas other moves let rhydon chunk me, than I should click blizz 100% of the time (unless I'm severely behind and need to claw my way back with lots of risky plays like this).
Optimal vs exploitive
While thinking about how to exploit X frequency of Y is an interesting discussion, it assumes you know what the ideal frequency of Y. We don't even know what the optimal frequency is (or frequencies, if there's multiple equilibrium strategies exist).
So the question of how to exploit something is harder than what the correct frequency is. And it's also less useful, whenever you exploit you are yourself being vulnerable to explotation, so it's only ever relevant in a tournament setting where you 'prep' against an opponent, if at all, and never relevant on ladder. The classical example of how exploiting means making yourself exploitable is RPS. You know you can exploit 100% rock by going 100% paper.
That said, I do think that thinking about how to exploit too low a freq and too high freq is relevant to finding out the ideal req, not sufficient by any means, but relevant. But maybe it's better to communicate our intuitions about what the ideal frequency is. If an AI specifically was designed or trained to play this gen, what would its Chansey freq would be? But before attempting such a hard question we have to lay the ground with some assumptions and theory.
Chansey value
First, let's talk about the value of each pokemon relative to the other 5 pokemon on the team, I think this gen has a very disparate value of between the 1st best mon on a team, and the last best mon on a team, unlike other gens. We can argue about why this is, maybe because there's less mons, maybe because there's no active banning system, but if we take the trouble of giving each mon a relative value, like we do in chess (Pawn=1, Knight=3, Bishop=3.3, Rook=5, etc...) most of us would argue that Chansey, Tauros and Lax would be above average, and Chansey would be first by a lot.
And this is what gets some people confused, a lower frequency doesn't mean a lower value necessarily. I think chansey has a value of 2, Tauros 1.3 and Snorlax 1.4. (Eggy 1.2, starmie 1.2, zam 1, zapdos 1, Rhydon 0.9, Gengar 0.8). It's hard to normalize the values, but I think of Zam as 1, it's goes at least even against almost anything and it takes between half an opposing mon and 1.5 opposing mons, rarely 0 or 2.
Disparity between value and frequency
Now this is what gets some people confused, a lower frequency doesn't mean a lower value necessarily. Dangerkid's saying
"It's not just about how strong the move/mon is, but how punishable it is."
I think he's correct. Disparity between value and frequency can only occur when it is punishable, like the example of modified RPS, the payoff of R win is higher, but it's still checked by paper, so the frequency becomes lower, and the frequency that goes up is actually the check/counter.
If the mon were so OP that it didn't have counters, then it would actually have a high frequency AND a high value, like MewTwo in ubers. Additionally I might add that the exploitability doesn't need to be of the individual mon, but of a team that contains it, and a team merely having a mon is much rare to exploit on its own.
Conclusions of naïve model
Taking these assumptions, that a chansey is worth 2 zams, we should assume that the ideal chansey frequency would be 100%. As it is not possible to exploit 100% chansey. Or, it would mean that 100% chansey is exploitable.
I think some dynamics may be missing from this model, team comp wise, Chansey competes with other mons for the special wall slot, so it's possible that other mons simply fulfill that role much better. Or it might be about synergies and team comps.
Synergies and team comps, specifically HO
We have to talk about HO, it's very hard to fit Chansey on HO, so we would see lower Chansey Freq because of it. And in those HO teams, the value of a chansey might not be the same as in a wall team, so a team that chooses HO may lose value on their not just by not selecting chansey, but by not having wall support in general (suppose for example that Chansey needs at least 1 other special wall to function effectively). Furthermore the value of chansey on the opposing team is greatly reduced as well, if team A has one starmie and a Chansey on their team, the opposing chansey is worth like 2 points, as it's a whole wincon. But if Team A has 6 mon that can boom or hyper beam chansey, then chansey is all of a sudden worth 1 point or less.
I personally think that 2 special walls and 4 offensive mons work fine, but that doesn't mean that 6 offensive mons work too.
Actually, 6 Boom/Hbeam HyperOffense is exactly a team crafted against chansey, if you replace one offensive mon with a special attacker that can't boom, like starmie, then chansey gets its value. The way to exploit 100% chansey, might be to slap in 6 offensive mons.
That said, I never thought of HO as a counter to Stall comp, it might be worth it to have a lot of matches be played with asymmetric rules, like one team has to use chansey, and the other has not to. And see what comes out, it should end up as stall vs HO, and we might get a veredict on whether one or the other has an advantage or it's drawish. Even if it's drawish, HO might fare well against other choices, which might explain low frequency, high value chansey.
. She definitely does the most heavy lifting compared to the other mons. And in a consistent way I might add 
I would disagree with Rhydon though
because he has both defensive and offenseive value that is very high on the sorts of teams that I use. But I know most people don't play like me so I don't know how that fits into the frequency/value argument because frequently Rhydon is more expendable on a team with Lax and Tauros on it.
