Should you tail picks from X?

Tailing a pick means copying a bet someone else posted - a slip screenshot on X, a play in a group chat, a capper's card. The interesting question is not whether the person posting it is sharp. It is whether the price still on the board is a good price, and that part is checkable in about ten seconds.

Written by Jesse, NegativeEV. Last updated 12 August 2026.

What kind of bet gets posted in the first place?

Bets get shared because they are exciting: long prices, big parlays, the ticket worth screenshotting. That is also the worst-priced corner of the board. Across 23,059 priced MLB selections from 252 games, both sides, at the best price available anywhere, graded against the simulations rather than against real outcomes, 55.5% came out negative expected value. In the +400-or-longer band it was 71%. From +200 to +399, 69.7%. The shape that makes a pick worth posting is the shape that grades worst.

That is a measurement of the board those picks are drawn from, not of the picks themselves. Nobody has graded every pick posted on X. But a bet has to come from somewhere, and the corner of the board that produces screenshots is the corner where seven in ten selections are priced against the bettor.

The price on the post is not the price left for you

Expected value is a function of the number on the slip. A pick that was fair at +260 an hour ago is a different bet at +215 now, and a post that got attention is exactly the kind that moves the number. Tailing late is not the same bet, even when it is the same team, the same player and the same market.

Does a good record make the next pick a good bet?

A record answers a question about the past. Each new bet still has one price attached to it, and that price is either better or worse than the real chance of the thing happening. Those are separate questions, and only the second one can be settled before the game starts.

What does this page not establish?

Anything about the actual picks people post.

The 55.5% and the 71% describe the board those picks are drawn from, graded against our simulations rather than against what later happened. Nobody has graded every pick posted on X, and this page has not graded yours. The figures are also priced at the best number available anywhere, so a bettor locked into one book is looking at worse odds than the ones measured here. And none of it reads a single post: a poster, a record, or tonight's price can each be better or worse than the band they sit in. What stands is the direction, that the shareable corner of the board is priced against the bettor, not a verdict on any one pick.

How can a pick be checked before tailing it?

NegativeEV simulates each game play by play, thousands of times, and counts how often the bet actually wins across those simulations. That rate is compared with what the posted price implies, and the bet comes back graded as an expected value.

Free, no account needed, covering MLB, WNBA, PGA, and ATP. Check the pick before tailing it: https://negativeev.com

More on the number: https://negativeev.com/about/expected-value

More on the bot: https://negativeev.com/about/x-bot

Related reads: what happens when a bet is checked and how much of a sportsbook board is negative EV.