Best Dota 2 Player Props Today, November 4, 2026 (Data-Backed)

Today’s dota 2 player props board is built for November 4, 2026, and this page gets refreshed every day instead of republished as a new post. So the numbers below always describe the current series, not something written last week. Everything here comes from one match, GamerLegion versus OG, and every pick is the over side settling inside a map window.

Seven props made the cut. Five of them cleared at 70% or better across the last 10 series, but they all share the same game script, which changes how you should stake them. More on that below.

The short version: the death overs look sturdier than the kill overs, and one Natsumi- line is fading fast.

Player death props for GamerLegion vs OG

Death props settle on a single map. They’re less about who wins and more about how messy the map gets, which is why the same series can produce four of them without the picks fighting each other.

Today’s picks at a glance

Hit rate: how often the player cleared this line over his last ten games or as many as he has played (in brackets); L5: the last five. Each pick’s card and write-up follow below.

Natsumi- OVER 3.5 map 2 deaths (-106 at thunderpick)

Natsumi- has gone over 3.5 deaths on map 2 in 8 of his last 10 series, and the last five are a clean 5/5. Against GamerLegion he’s 3/3, so the head-to-head doesn’t fight the form. The recent-form gap sits at +32.8% and the price is -106, close to a coin flip. That gap is measured against a de-vigged fair price on a 10-series sample, which inflates it, so read it as a reason to shop the number rather than a promise.

Correlation: Natsumi- also shows up on the map 1 kills prop further down this page. Those are the same player in the same series, so stacking them doubles one risk instead of spreading it.

TOPSON OVER 4.5 map 1 deaths (-123 at thunderpick)

TOPSON is 8/10 on this line, and the last five hold at 4/5, so the recent trend isn’t quietly disagreeing with the headline number. Map 1 props reward early chaos, and his sample says he finds it. The recent-form gap is +29.3% with the price at -123, meaning you’re paying real juice for the privilege. Ten series is a thin sample, so line shop before you commit.

Skem OVER 6.5 map 2 deaths (-135 at thunderpick)

Skem’s 7/10 rate actually improves in the short window, at 4/5 over the last five series. He’s cleared 6.5 deaths in all three meetings with GamerLegion, which is the kind of head-to-head that supports the number rather than contradicting it. The recent-form gap is +17.2%, the smallest of the death props on this page, and the price has moved out to -135. You’re laying more juice for a lower gap.

OneJey OVER 7.5 map 2 deaths (-141 at thunderpick)

OneJey sits at 7/10 with 4/5 in the last five and 3/3 against GamerLegion, so nothing in the data pushes back. The catch is the price. A +16.3% recent-form gap at -141 means the market already respects the form, and this is the most expensive prop on the board. It’s the thinnest number here, even with the clean head-to-head.

Player kill props for the same series

Kills swing harder than deaths. Map length, role and whether a team is playing from behind all move the number, which is why the trend split matters more here.

Yuma OVER 6.5 map 2 kills (-130 at thunderpick)

Yuma is 8/10 on map 2 kills and 5/5 over the last five series, so the long sample and the short one agree. That’s not always true on this board. The recent-form gap is +28.1% against the de-vigged line, priced -130. Ten series is the entire argument, so treat the hit rate as a form read and not a projection.

Natsumi- OVER 6.5 map 1 kills (-108 at thunderpick)

Here’s where the numbers argue with themselves. Natsumi- is 6/10 across the last 10 series but just 1/5 across the last five, and he’s hit this line once in three meetings with GamerLegion (33.3%). That’s a fading profile, and saying so is more useful than hiding it. The recent-form gap is +12.4%, the smallest figure on the page, and -108 is the only thing keeping it interesting.

Correlation reminder: this is the same player as the map 2 deaths over. A quiet map 1 and an early map 2 death can happen in the same series, and both props move together.

Team market: total kills for GamerLegion vs OG

GamerLegion vs OG OVER 53.5 total kills per game (-119 at thunderpick)

Both teams land over this number in exactly 50% of their last 10 games, so the combined rate is 50% and neither side is carrying it. The case for the over sits in the raw averages: GamerLegion at 54.9 and OG at 55.1 kills per game over the same window, both clear of the 53.5 line. At -119 the ask is modest for two baselines sitting a kill and a half above the number. It’s a fixture-level market, not a player’s line, and it settles on the series.

Correlation: these props share one game

Every player prop on this page comes from the same GamerLegion vs OG series, and picks 1 and 9 are the same person. That matters for staking, not for the analysis.

• A slow, low-kill series can sink the kill overs and the team total at the same time.
• A scrappy map 2 helps most of the death props but doesn’t rescue the kill overs.
• Natsumi- map 2 deaths and Natsumi- map 1 kills are one position, not two.

Stacking correlated props multiplies the same risk. It doesn’t diversify it. If you’re building a slip, treat the whole player group as a single position rather than seven independent bets.

How StatsBench builds today’s Dota 2 player props

Hit rates come from real game logs. For each player we take the last 10 series and count how often he cleared the posted line inside that map window. Map-window props like maps 1-2 or maps 1-3 exclude best-of-one series by design, since there’s no second map to settle against.

The recent-form gap is the last-10 over rate minus the de-vigged market-implied probability. It’s a form signal only. It is not a projection, not expected value, and not a return figure. A 10-series sample overstates a big gap, so a wide number is a reason to check the current line across books, not a green light. Our validated long-run advantages are much smaller, in the single digits.

Prices come from StatsBench’s own odds feed across 30+ sportsbooks, DFS apps and exchanges, updated through the day. Esports markets on StatsBench are research: hit rates with the sample attached, both sides of a duel market, and the posted price. Nothing more.

Bet responsibly

This is research and entertainment, not income. Every number on this page is a past result and none of it guarantees a future one. Long-run variance is real, and a 10-series sample can look very smart for a week and very dumb the next.

Bet within your means, only with money you can lose, and take a break when it stops being fun. 21+. If gambling stops feeling optional, call 1-800-GAMBLER.

Frequently Asked Questions

What does map 2 deaths over 3.5 mean in Dota 2?

It’s a prop on how many times a specific player dies during map 2 of a best-of series. Over 3.5 means he needs at least 4 deaths on that map to cash. The bet settles on that single map, not the full series.

How is the recent-form gap calculated?

It’s the player’s last-10 over rate minus the de-vigged market-implied probability on the same line. It measures recent form against a fair price, and it’s not a projection or an expected return. A big gap off only 10 series tends to overstate itself.

Why does today’s page only cover GamerLegion vs OG?

The snapshot that day surfaced the deepest prop board on that one series, so every pick comes from the same match. That also means the picks are correlated, and stacking them concentrates risk instead of spreading it.

Does a 5/5 last-five hit rate mean a prop will cash?

No. Five series is a tiny sample and it can flip fast, which is why we compare it against the last 10. When the two disagree, like the Natsumi- map 1 kills line, the short trend is often the more useful warning.

Are these picks or research?

Research. StatsBench publishes hit rates, game logs and posted prices, and you decide what to do with them. There are no guarantees here, and every edge is a long-run idea rather than a sure thing.