Race-Time Optimization — Verified Records and the 5-Second Soft Cap

🔬 Sourcing. Every claim on this page is anchored to a public YouTube playthrough with a timestamped link. Statements that cannot be traced to a primary video clip have been removed from this guide rather than hedged. The four cited videos are HeTheLyon's "Only Humans" run, MoreCaseOh's "Fastest Horse" run, Cheru's "100% Solved" run, and bwhatsonn's "Best Genes" run — full URLs in Sources at the bottom.

The skeleton question of Horsey Game racing is: how fast can a creature actually go? This guide is the answer with receipts.

The verified race-time progression

HeTheLyon's optimization run starts with an unmodified human and reaches the apparent floor through iterative DNA editing. Each time below is observed on screen and called out by the creator:

| Iteration | Time | Source | |---:|---:|---| | Baseline (unmodified human) | ~60s | @ 0:57 | | Sub-30 target stated | 30s goal | @ 3:10 | | First optimized time | 27s | @ 3:41 | | Mid optimization | 22s | @ 8:20 | | | 20s | @ 8:23 | | | 19s | @ 8:28 | | Pogo-leg breakthrough | "actually won" | @ 11:47 | | | 12s | @ 12:08 | | | 9s | @ 16:29 | | | 7.9s | @ 16:40 | | RAM-upgraded simulator | 6.9s | @ 18:37 | | | 5.2s | @ 20:19 | | Apparent floor | 5.0s | @ 20:27 |

This is one of the cleanest optimization curves we have video evidence for in any breeding sim. A 12× compression from baseline to floor, with the curve flattening sharply below 8 seconds.

The 5-second soft cap

HeTheLyon @ 20:30 verbatim:

"I swear, there's got to be like a soft cap. That lets it to where they can't go below 5."

He has not personally beaten 5.0 seconds at the time of recording, despite stating earlier @ 16:45 that he has "never gotten below five seconds" across his sessions. Treat 5.0s as the public record floor, with one important asterisk below.

The hot-sauce sub-5 speculation

HeTheLyon @ 22:21:

"The hot sauce might have made them faster. That was faster than 5 seconds."

This is the only public claim of a sub-5-second time we've been able to source. He flags it as uncertain in the same sentence — the camera doesn't clearly show the time. We're tracking this as a community open question. If you have a clear screenshot of a sub-5 time, please submit via /contact — it would be the first hard confirmation.

Genetic interventions that demonstrably worked

These are mechanics HeTheLyon visibly applied during the optimization run:

Empty the DNA vat between iterations (@ 6:08) — "There isn't shitty genes clogging it up." Verified on screen as a step that improved subsequent times.
Plutonium feeding (@ 7:11) — described by the creator as "feed them a little bit of plutonium to speed them up." Visible side-effect: the creature grows hooves on its feet.
Pogo front leg (@ 11:47) — "Don't know why he has a pogo for his fucking front leg. But hey. I think it's working. He actually won." Specific morphology that translates into wins.
RAM upgrade for the horse simulator (@ 19:56) — "With the RAM, we can upgrade our horse simulator. It'll make even faster horses." Game offers a hardware-tier upgrade path that improves optimization output.

What this batch could not verify

Several frequently-claimed mechanics did not appear in the four cited transcripts. We've listed them here so you know what's still open:

The "5–15% race-outcome delta" rule of thumb
A pace slider with 50/75/90/100/110% settings
A Feed Bonus / Warm-Up / Track Affinity buff stack worth +12% Speed
Rain -3%, Heat +20% Stamina drain, Cold 1.5s start-line delay
Track-specific stat thresholds (Meadow Sprint Speed ≥85, Mountain Pass Stamina ≥90, etc.)
"Wet-weather Hoof Structure (H9=C)" or any allele-letter-to-trait mapping

These weren't in the four videos. They may exist; we don't know. Until a save-file extraction lands or a video shows the mechanic on screen, they're not on this page.

Practical takeaways

Your baseline is ~60 seconds. A fresh creature is slow. The optimization process compresses 12× to 5.0s.
Empty the vat between optimization iterations. Cumulative DNA pollution slows you down.
Plutonium and pogo-leg morphology are documented speed paths. Not pretty, but they win.
RAM upgrade unlocks a faster simulator output. It's a real progression gate.
Don't trust race-time claims that say "sub-5 seconds" without a screenshot. The current public floor is 5.0s — sub-5 has been speculated about but never visibly demonstrated on a clean clip.

What's next

How DNA Actually Works in Horsey Game — the model behind the optimizations
DNA Database — the editor where iterations happen
Opening Speedrun — community chain that gets you to the optimization phase fast
Champion-Tier Build — different optimization target (championship win, not race time)

Sources

HeTheLyon — "I Beat Horsey Game With Only Humans". https://www.youtube.com/watch?v=iSYZvzz--NQ
MoreCaseOh — "We Created The Fastest Horse Of All Time…". https://www.youtube.com/watch?v=I4f8FkUT8rc
Cheru — "I 100% Solved Horse Racing — CRISPR and a Broken Moral Compass". https://www.youtube.com/watch?v=CoiOVMPzUWM
bwhatsonn — "Mixing THE BEST Horse Genes". https://www.youtube.com/watch?v=A9dgl_TSBRA

Last updated: 2026-05-09. All race-time claims on this page have a primary-source video timestamp.

Last verified: 2026-05-09 · Game version: 1.0.4 · See /sources for citation list.