here's the slippage over the 3 exchanges I mainly trade now, hyperliquid, extended and .

hyperliquid and extended's median are very close together.
it may be dumb but I thought this was because of the fact that extended only accepts "limit-type" orders. but hyperliquid also has the same mechanism on market_open which the SDK converts them to a limit IOC with a 5% slippage buffer (i should know this but i didnt tbh).
been just awful.
what's interesting is that I trade relatively similar sizes on extended as I do on hyperliquid, so it's not even a size thing.
it's also interesting if we look at it from a size perspective (I've hidden the $ size per order on purpose).
the first 4 buckets, extended's average slippage remains quite stable, while on the last largest 20% of orders, it triples, while the median stays relatively stagnant. the degradation should come from a minority of ugly fills, not an uniform degradation, if that makes sense. this is because for a long time I was trading there pairs that I shouldn't have, and maybe be spilling over into here.
it just goes to show that at the tails of size extended degrades way less gently than hyperliquid's.
this is obvious stuff right, hyperliquid is much more mature, but its interesting to me to see it on data.
's slippage is worst on the smallest orders, which is interesting. this is probably a long tail pairs problem, that are more volatile, and get less size. on stuff I trade more size, eg btc, eth, hype, it has a thicker book, and the larger size orders get better execution. in any case I only trade a small % of my book there for now. I need to fix the execution before I trade there any relevant size.
next month I am going to work on all orders being limit across exchanges, and this should improve. especially for what i do it doesnt make sense to be crossing the spread like an ape, i dont need speed.
different strats will require different treatment but yeh.