CS2 Kill Props: Over or Under for Stars?

A CS2 kill prop is a two-sided market, and the line is the whole argument. One player’s map 1 kill total gets an over and an under, each with its own price. The hit rate tells you how often the over cleared in his last ten maps. It does not tell you whether the price is right.

That gap between the record and the price is where most CS2 kill props get misread. Below is a real set of map 1 lines from G2 vs Spirit at BLAST Rivals Fall 2026, pulled from StatsBench, plus a method for weighing both sides before you take one.

What a CS2 map 1 kill prop actually prices

A map 1 kill line prices one player in one map. Not the series, not the tournament. That matters, because map one is often the least predictable map of a series: both teams have prepared it, and the veto can leave a layout neither side loves.

Kills are also a noisy stat. A player who dies early three rounds running can still finish on 20 if his team forces and he cleans up exit frags. Overtime tacks on five to ten kills by itself. Eco rounds swing totals hard, and usually toward the team that is winning.

So the book does not post a number it thinks is fair and walk away. It posts a number and a price, and the price is where the argument lives.

Why the star’s kill line is usually the hardest to clear

Take donk. StatsBench listed his map 1 kills line at 19.5 against G2, with the over at 3/10 in his last ten maps and a best price of -133. Ugly record, yet the price still says the over is the likelier side.

That is not a contradiction. A 19.5 line is a high bar, and the market knows donk takes more duels than anyone else on Spirit. Those ten maps were also ten specific opponents on specific maps, and none of them asked him for 20 kills.

Now put magixx next to him: 12.5 kills, 6/10, best price -111. Lower line, better record, almost the same price. If you only looked at the percentages, you would conclude magixx is the stronger play. You would be answering two different questions.

One match, nine lines, one lesson

Here is a slice of what StatsBench listed for G2 vs Spirit in that November 2026 BLAST Rivals Fall slate. Each row is the over, its last-ten record, and the best price available on that side when the board was pulled.

Player	Market	Line	Last 10	Best price
Nertz (G2)	Map 1 kills	14.5	7/10 (70%)	-109
Nertz (G2)	Map 1 headshots	7.5	7/10 (70%)	-139
magixx (Spirit)	Map 1 kills	12.5	6/10 (60%)	-111
r1nkle (G2)	Map 1 kills	11.5	6/10 (60%)	-128
zont1x (Spirit)	Map 1 kills	13.5	5/10 (50%)	+100
HeavyGod (G2)	Map 1 kills	13.5	5/10 (50%)	-133
tN1R (Spirit)	Map 1 kills	14.5	3/10 (30%)	-105
donk (Spirit)	Map 1 kills	19.5	3/10 (30%)	-133
zont1x (Spirit)	Map 1 headshots	5.5	3/10 (30%)	+100

The lesson sits inside the table: a hit rate only means something next to the line it was set at. Seven of ten at 14.5 asks for a different map than three of ten at 19.5.

Headshot props ask a different question

Headshot markets look like kill markets and are not. A kill counts once. A headshot only counts if the final bullet lands high, which means role and weapon matter as much as volume.

Nertz at 7.5 headshots, 7/10, priced -139 is a player the data says finishes high consistently. zont1x at 5.5 headshots, 3/10, priced +100 is a much lower bar with a much worse record, and the market still only makes it a coin flip. Read that as the book telling you it does not trust the three recent misses either.

Before either number means anything, check how the player is used. Riflers who hold tight angles build headshot counts differently than an AWPer, whose body shots never appear in this column at all.

The math that makes two lines comparable

zont1x at 13.5 kills, 5/10, priced +100. HeavyGod at 13.5 kills, 5/10, priced -133. Same line, same record, 100 cents of price apart. Convert both to implied probability and the gap stops being abstract.

• +100 implies 50.0%
• -105 implies 51.2%
• -109 implies 52.2%
• -133 implies 57.1%

The implied probability calculator does that conversion in one step. Once both prices are percentages, the record stops being a vibe and becomes something you can argue with.

Then there is the hold. The over and under on the same prop usually add up to more than 100%, and that extra is the book’s margin. Strip it out and you get a cleaner read on what the market actually believes. The de-vig calculator handles the pair.

Ten maps is a thin sample, and CS2 punishes thin samples

A 7/10 record has a lot of room around it. Ten maps means ten vetoes, ten opponents, and probably two or three different map types. Change the economy or the side you start on and the same player’s kill total moves by five.

Treat the record as a hint about role and volume, not a probability. Nertz at 7/10 on 14.5 kills tells you he is in a lot of rounds. It does not tell you he clears 14.5 seven times out of ten against this specific five. There is a longer version of that argument in sample size in betting.

What a hit rate cannot see: pistols, map pool, overtime

Pistol rounds decide the first three rounds of each half. A team that loses both pistols spends long stretches on eco, and eco rounds distort kill counts badly. The winning side feeds on them. The losing side’s star often does not.

The free CS2 Pistol Lab shows both teams’ pistol rates side by side, which is the cheapest way to sanity-check a kill line before you take it. If one team wins pistols at a high clip, the other team’s kill overs get harder in a way a last-ten record will never show you.

Map pool does the same work. A long, angle-heavy map produces more engagements than a fast one, and overtime is always live in a close map one. Roster news and map history sit on Liquipedia’s Counter-Strike pages, and HLTV carries the per-map scorelines you need to see which layouts actually produce kills for each player.

A routine you can repeat

1. Pull the line, record and best price from the CS2 prop finder. Write the record as a fraction, not a percentage.
2. Convert the price with the implied probability calculator, then compare it to the fraction.
3. Check the other side and de-vig the pair. A wide hold means the market is not confident either.
4. Look at the map and the pistol numbers in the Pistol Lab.
5. Shop the same player’s line across books. The best price on the board is not always on the book you already use.

For a worked example on a single match, the Spirit vs G2 preview walks the same lines player by player. If you want the grading side, reading kill lines covers how a map 1 total gets settled.

Every CS2 match on the slate, with lines and records, sits at the esports hub. For the daily list, the CS2 picks page posts the numbers each morning.

One last thing worth saying out loud: a kill prop is research, not a promise. A ten-map record misses more than it catches, and variance in a best-of-three is real. Bet within your means, 21+, and keep your own sample honest in the Bet Tracker.

Frequently Asked Questions

What is a CS2 kill prop?

It is a bet on one player’s kill total in a set map or match. Most books split it by map, so a map 1 kill line only counts kills on the first map of the series, overtime included. The over and the under each carry their own price.

How should you use a 10-map hit rate on a kill prop?

As context, not as a probability. Note the line the record was set against, convert the price to implied probability, and compare the two. A rate from a 19.5 line says very little about how a 14.5 line should be priced.

Does the map matter for CS2 kill props?

A lot. Longer, angle-heavy maps produce more engagements, and the veto decides which layout you get. Pistol rounds and overtime also swing kill totals by several either way, and neither shows up in a last-ten percentage.

Should you take the over or the under on a star player?

Neither by default. High lines on high-volume players are priced for that volume, so the record can look bad while the over stays the likelier side. The decision comes from the line height, the price and the matchup, not from the player’s reputation.

Where can you find CS2 kill lines with hit rates?

The StatsBench CS2 prop finder lists map 1 lines with the last-ten record and the best price across books, DFS apps and exchanges. The free CS2 Pistol Lab adds both teams’ pistol rates, which is the fastest context check on a kill line.