Check the categories now ;)

I don't think this will be necessary on unseeded runs for a while, but for a seeded runs the gap in input devices is huge

what a seed, and what a run!

Holy Canolli

A few updates on this:

• the use of symbols was unintended and got patched, lowering the number of possible seeds back to 35^8

• The seed searcher is much faster; on my RTX 3080 it can search most things in about a day, which some searches being faster depending on the speed of certain calculations

• For those not in the Discord, I ran a seed search today for SS skips. I’d recommend checking out the world record to see the new seed I found.

For seeded runs, touchscreen is pretty much unbeatable

I was able to get multiple sub-1:20 runs on the new seed I found while runners using other methods have runs around a 1:25

a) Unfortunately as of now I can’t share a working copy. The seed finder is currently in a phase where changing the search criteria requires changing lines of code, and for the demo version it won’t be open-source. There is a UI in progress, and when that is finished I can send you a copy of the program.

b) Agreed, although the number of seeds is actually 2.25-4.34 trillion (the characters allowed in seeds are a bit complicated, as the number 0 is converted to the letter O and a few symbols (.^$ if I recall) are allowed in the seed input, sometimes only by pasting, leading to between 35^8 and 38^8 depending on what you want to use). For comparison, the number of seeds the searcher can search in a reasonable timeframe is in the range of billions, and with a few weeks of time it could likely search the full range of seeds.

I’ll be sure to upload any findings that can improve the WR. As a fun fact, the current WR seed was found with an older version of the seed finder (which explains why it has periods in it).

Besides speedrunning Balatro, I’m also working on a seed searcher. A seed with these exact conditions probably can’t be found, but I can find some pretty speedy seeds for a run where something like high card isn’t used and the other restrictions are more lenient.

Here are the splits from my 5:12. My sum of best segments is a 4:42. As for strategies, honestly the biggest difference I've thought of is using high card instead of pair, which could save time due to being easier and faster to play, with the downside of scoring fewer points. It's what the current SS skipless record uses as well.