Jetski_

Building the most accurate competitive modelling system for CS2

[Jetski_ @TimoThornton1]

@LuzRaposa Donk on another planet. Absolute madness.

[Jetski_ @TimoThornton1]

@Mauisnake I think you're right, but I think we both hope he proves us wrong.

[Jetski_ @TimoThornton1]

@karriganCSGO Please never retire

[Jetski_ @TimoThornton1]

@LuzRaposa Confirmed eye test. I would love to know if this was the case 4+ years ago.

[Jetski_ @TimoThornton1]

You don't even need to look at the stats. Ropz is the most winningest player OAT. He's an overpowered asset to any team. I find it weird he's never in the GOAT conversation. Certainly the best lurker there ever was. IEM Rio puts him amongst the all time greats.

[Jetski_ @TimoThornton1]

@LuzRaposa This is so interesting, really sets him apart

[Jetski_ @TimoThornton1]

@LuzRaposa He has a 33% T side opening duel win rate in big tournaments. CT side better at 49%. His overall Opening is only 3% better than Karrigan (40% vs 37%).

Rain still has an opening of 49%.

jcobbb does though, marginally, Trade kill the highest in Faze though.

Still, pretty poor

[Jetski_ @TimoThornton1]

CS2 teams scout by K/D and HLTV rating.
Neither metric knows what role you play.
We built JumpShot to measure structural contribution instead.

https://t.co/xF6yt9eTjX

[dAVE @CS2dAVE]

been working on a cs2 site with some tools i built - figured i'd share it!

https://t.co/dhlKvOEo71

it has a demo viewer where you can scrub through pro matches, and an AI match predictor

still early but check it out and lmk what you think i'd rly appreciate it 🙏

[Jetski_ @TimoThornton1]

@CS2dAVE Awesome! Please check your DM's

[Jetski_ @TimoThornton1]

> Be Jame
> Refuse to change style
> Build team around it
> Recruit aggressively
> Lowest pistol win rate of top teams (37%)
> need economy boys
> Win 14% of full-eco's (Elite)
> ok, but deny their eco's boys
> Win 93% of anti-eco's
> Win 72% post-plants
> Need entries
> ok fine, I'll do it (T 62%, CT 65%)
> Make finals
> Profit

[Jetski_ @TimoThornton1]

@Magestia_dev @Ozzny_CS2 Ropz joined faze with Karrigan and they won a major and the grand slam. Him leaving faze to join vitality has dropped faze out of the top 10 and made vitality quite possibly the greatest team of all time. On donk, yes, but he has the potential to be there in a few years

[Jetski_ @TimoThornton1]

The JameTime Paradox (?)

Cluj-Napoca was an eye opener

Jame now sits at #7 in our global rankings, above other elite AWPers like Molodoy and torzsi. His four teammates occupy positions #100 through #132. No other player in the global top 10 has a supporting cast placed this far back. On paper, this looks like structural fragility.

PARIVISION are, obviously, not trying to be stars. They are maintaining the conditions that keep jame armed. Tight economic discipline, utility execution, and map picks engineered around CT-side frameworks are their contribution to the system. When that system works, jame has a full AWP in a full-buy round against opponents who cannot afford the same. At that point, our model shows what happens: 93.2% of anti-eco rounds converted, and 71.7% of post-plant situations won. These are match-deciding numbers.

The CT/T asymmetry confirms the structure. PARIVISION win 51.4% of CT rounds against 44.4% on T-side. Their map picks reflect this consistently: Inferno, Dust2, Anubis, and Train all appear as CT-favoured selections in the data. This is not accidental. Their map pool is built to maximise the number of rounds played from CT, where a correctly positioned Jame is at its most dangerous.

The failure point is also visible. PARIVISION win just 37.2% of pistol rounds, the lowest rate in our dataset. Pistols compress the economy back toward zero, bypassing the full-buy framework entirely. When they lose a pistol, they enter a cascading economic state where jame's AWP cannot be funded. The eco round data is the tell: despite winning 14.3% of eco rounds (highest in the dataset), even that reflects careful resource management under adverse conditions rather than aggressive firepower.

This is how PARIVISION reaches finals. Not by fielding five individually dominant players, but by compressing matches into the scenarios where their system is optimised. The question opponents face is not whether they can beat jame individually. It is whether they can force enough non-standard economic states to break the structure before it closes the map out. When they cannot, PARIVISION win. The data says they do that more often than the individual rankings would suggest possible.

[Jetski_ @TimoThornton1]

@PimpWinneche Far more depth than Astralis had. Cannot be underestimated how good the structure of this team is, on par with prime Astralis. Plus ZywOo is consistently indomitable.

[Jetski_ @TimoThornton1]

@blinkdotnew Blink

[Jetski_ @TimoThornton1]

@antmillionsbot 👀👀

[Jetski_ @TimoThornton1]

@ashhhcs @tvoy_molodoy @ashhhcs - I've been building an CS analytics model for the last year, would love your opinion if you've got some time, DM me.

[Jetski_ @TimoThornton1]

@tomeromrix @base44 Yes please

[Jetski_ @TimoThornton1]

@pastalj @JustMani @Springboks pool E kinda

[Jetski_ @TimoThornton1]

@learn2vibe @voicenotesai "Senior vibe coder"