CS2 prop picks today: what the numbers say on August 23, 2026
This page refreshes every single day, and today’s board is dated August 23, 2026. Below are eight data-backed CS2 prop picks today, pulled straight from recent series logs and priced against live sportsbook numbers. Every line, hit rate and price you see is exactly what the StatsBench feed shows, nothing rounded up for effect.
Fair warning before you scroll: almost the entire card comes from one match, FUT vs Spirit. That’s a correlation problem, and I’ll flag it again next to the picks.
How to read the hit rates on this page
A CS2 prop settles over a map window inside a best-of-N series. So a “90% hit rate” means the player cleared that line in 9 of their last 10 series, not 10 individual maps. Map 1 props always have a map to settle on. Map-window props (maps 1-2, maps 1-3) drop best-of-1 series from the sample by design.
The “recent-form gap” is the last-10 over rate minus the de-vigged market-implied probability. It’s a form signal, not a profit projection. Ten series is a small sample, so a huge gap is a reason to shop the price, not a reason to fire blind.
Krabeni OVER 13.5 map 1 kills vs Spirit
Krabeni has gone over this number in 90% of recent series (9/10), and the last five are a clean 100%. That’s a form curve pointing in one direction, with an L10-vs-market gap of +37.3% against the de-vigged price of -133 at thunderpick. The two windows agree here, which matters: a hot L5 sitting on top of a strong L10 is a real trend, not a blip.
One caution. A gap that large off ten series almost always overstates the true edge, so treat -133 as a number to shop rather than a mispricing you must have. Verify the current line before you commit.
salazar OVER 11.5 map 1 kills at Lavked
The only pick on this card outside the FUT vs Spirit match. salazar has cleared 11.5 in 90% of recent series (9/10) with an 80% mark over his last five, so the sample stays consistent across both windows. The recent-form gap sits at +36.1% versus the -143 price at thunderpick.
An 11.5 map 1 kill line is a low bar for a starter playing full maps, which is part of why this rate holds up. Still, -143 is real juice. You’re paying for the safety, and the small sample means the true clearance rate is likely lower than 90%.
dem0n OVER 14.5 map 1 kills vs Spirit
dem0n’s over rate is 70% across recent series (7/10), and his last five sit higher at 80%. Trend and baseline agree, which is what you want before you take a 14.5 line at -118 on thunderpick. The recent-form gap is +20%.
This is the highest FUT line on the card outside of donk’s, so the price reflects respect for his role. He’s the more reliable of the two big FUT numbers based on the L5 versus L10 alignment.
magixx OVER 11.5 map 1 kills at FUT
Here’s the one to be honest about. magixx shows 70% over recent series (7/10), but his last five are only 40%. That’s a fading player, and the headline number is being propped up by older results. The recent-form gap of +16.2% at -139 leans almost entirely on that stale half of the sample.
I’d still list it because 11.5 is a soft line for a Spirit starter, but this is the weakest hold on the board. Bet it smaller than the rest, or pass it.
donk OVER 20.5 map 1 kills at FUT
donk carries the highest line here at 20.5 map 1 kills, and he’s cleared it in 60% of recent series (6/10). The last five jump to 80%, so the recent direction is up while the ten-series baseline stays modest. The price is the cheapest on the card at -105 with a +12.7% recent-form gap.
Getting near even money on a star fragger whose last five say 80% is the best price-to-form combination on this page. High lines carry more variance, though. One quiet map and it’s gone.
dziugss OVER 13.5 map 1 kills vs Spirit
dziugss sits at 60% over recent series (6/10) with the last five also at 60%. No trend either way, just a steady rate that beats the -109 price by a +11.8% recent-form gap. Stability like that is actually useful when you’re sizing bets.
Note the overlap: this pick and the dziugss headshots prop below are the same player. Take one or the other, or accept that you’re doubling up on a single performance.
xfl0ud OVER 14.5 map 1 kills vs Spirit
xfl0ud has gone over 14.5 in 60% of recent series (6/10), with the last five at 80%. The short window is running ahead of the long one, which reads as a player growing into a bigger fragging role. The recent-form gap is +11.4% at -111 on thunderpick.
That gap is modest by this page’s standards, and honestly that’s a point in its favor. Smaller gaps off real form tend to hold up better than 35% outliers.
cmtry OVER 13.5 map 1 kills vs Spirit
cmtry is the sharpest split on the card: 60% across recent series (6/10), but 100% over the last five. Every one of his recent outings cleared 13.5. The recent-form gap of +10% at -118 does not fully capture that swing, because the ten-series figure drags it down.
Five series is a tiny window, so don’t treat a perfect L5 as proof of anything. It’s a signal that his role has changed recently, and the book’s line hasn’t fully caught up.
dziugss OVER 7.5 map 1 headshots vs Spirit
Same player, different stat. dziugss has cleared 7.5 map 1 headshots in 60% of recent series (6/10), with the last five also at 60%. Consistent both ways, +10% recent-form gap, priced -118 at stake. That’s a different book than most of this card, so line-shopping matters here.
Headshot props track kill volume closely, which is exactly why pairing this with his kills over is a stack, not a hedge.
FUT vs Spirit OVER 21.5 map round total
The fixture market backs up the player card. FUT have gone over 21.5 rounds in 60% of their last 10 maps and Spirit in 70%, for a combined 65%. Spirit are carrying that number, though the averages are close: FUT 23.4 rounds, Spirit 22.1 over the last ten. Priced at -109 with a listed EV of +25%.
Both sides average comfortably above the line, which is the cleaner read than either team’s individual rate. Longer maps also feed the kill props above, so this is the natural anchor of the card.
Correlation warning: read this before you build a slip
Eight of these ten rows come from the same match. Picks 1, 3, 4, 5, 6, 7, 8 and 9 are all FUT vs Spirit, which means one game script drives all of them. A fast 13-2 stomp kills most of the kill overs and the round total at the same time.
Two specific overlaps to name:
- Picks 6 and 9 are the same player (dziugss kills and headshots). One performance settles both.
- Picks 1, 3, 6, 7 and 8 are all FUT teammates. Kills are a shared pie, so they compete with each other.
Stacking correlated props multiplies the same risk. It does not diversify it. Pick 2 (salazar) is the only genuinely independent play on this page.
How StatsBench builds these hit rates and gaps
The method is deliberately boring. StatsBench pulls real recent series logs for each player, checks how often they cleared the exact line a book is offering right now, and reports that as the hit rate over the last 10 series (plus a last-5 view for trend).
The gap number is the last-10 over rate minus the de-vigged market-implied probability from the posted odds. De-vigging strips the book’s juice out so you’re comparing a real rate to a fair price, not to the sticker. That’s a recent-form signal, not expected value and not a projection.
Prices come from roughly 60 tracked sportsbooks, and the +EV feed recomputes fair prices about every minute. Lines move. Always confirm the number before you bet it.
A note on sizing and variance
A 60% hit rate over ten series still loses four times in ten. Ten series is a small sample by any statistical standard, so large gaps overstate the true edge. Validated long-run edges are far smaller, usually single digits.
Bet flat, shop every line, and judge yourself on process rather than one night’s result.
Responsible gambling
This page is research, not advice, and nothing here is a guarantee. Betting is 21+ and legality varies by where you live. Only stake money you can afford to lose, expect long stretches of variance in both directions, and if it stops being fun, stop. Help is available at 1-800-GAMBLER.
Frequently Asked Questions
What does a CS2 prop hit rate actually measure?
It’s how often a player cleared that exact line across their recent series, not individual maps. One series contains several maps, so a 60% rate means they went over in 6 of their last 10 series. Map-window props like maps 1-2 exclude best-of-1 series from the sample by design.
How often does this page update?
Daily. The same URL refreshes with a new board every day rather than publishing a new post, so today’s version is dated August 23, 2026. Lines and prices move constantly, so always confirm the current number at your book.
What is the recent-form gap and is it the same as EV?
No. The recent-form gap is the last-10 over rate minus the de-vigged market-implied probability from the posted odds. It’s a form signal only, and a 10-series sample tends to overstate it. Treat a big gap as a reason to shop the line, not as expected value.
Why is stacking these picks risky?
Eight of today’s rows come from the same FUT vs Spirit match, and two of them are the same player (dziugss kills and headshots). Correlated props multiply one shared risk instead of spreading it. A one-sided stomp can sink the whole slip at once.
Should I bet a prop that hit 90% of the time?
Not automatically. A very high recent hit rate off a small sample can point to a stale line or a data quirk as easily as a real edge. Verify the current price first and size accordingly, because no hit rate guarantees the next result.