Run the same NFL spread through a devig calculator twice — once from a sharp book's prices, once from a soft book's — and you get two different fair lines for the same game. Neither run made an arithmetic mistake; they just disagreed about which prices count as evidence. The Portfolios rebuild on OddsShopper put that exact choice directly in your hands: you can price your bets against OS weights, the preset reference-book weighting OddsShopper maintains, or flip to Custom weights and decide yourself which books define fair. That turns a settings toggle into a real strategy question: whose fair line should you trust, ours or yours?
That question is what this guide answers. We will walk the math of a devig once, look at what presets and custom devigging each get you, and finish with the one number in your bet history that settles the argument better than any opinion. If you are still getting comfortable with expected value or why the fair line is the benchmark for every bet, those primers are the foundation this piece builds on.
The Quick Answer
The preset is the right default: OS weights apply one consistent, market-tested view of which books matter to every bet, which keeps your fair lines comparable and your track record honest. Going custom is worth it only when you have a real, specific belief about which books are sharpest in your markets, and you are willing to grade that belief against the closing price over a real sample of tracked bets, hundreds rather than dozens. The worked vig math, the decision table, and the review-tab check that tells you whether your custom line is actually smarter are all below.
What A Devig Calculator Actually Does
A devig calculator strips the sportsbook's margin out of a market so you can see the fair, no-vig probability underneath. Take an NFL point spread with both sides at -110. Each -110 price implies a 52.4% chance of winning, so the two sides add up to 104.8%. Real probabilities have to sum to 100%, and that extra 4.8 points is the vig, the book's built-in margin. Remove it proportionally and each side is a fair 50%, which means fair odds of +100. The book sold you a coin flip at -110, and the devig showed you what the flip was actually worth.
| Side A | Side B | Total | |
|---|---|---|---|
| Posted Odds | -110 | -110 | |
| Implied Probability | 52.4% | 52.4% | 104.8% |
| Fair (No-Vig) Probability | 50% | 50% | 100% |
| Fair Odds | +100 | +100 |
That 4.8 points of margin is the tax you are trying to out-price on every bet, and it is why the fair line matters more than your opinion of the game. A bet is +EV when the price you got beats the fair line. You can run any single market through the free EV calculator, and the hold calculator shows you how much margin a book baked into any two-sided market before you bet it.
The one-market version is simple. The hard question hides one layer down: which prices do you feed the calculator in the first place? Devig a market from a sharp book's odds and you get one fair line; devig it from a soft book's odds and you get another. That choice of inputs, not the arithmetic, is the entire custom-versus-preset debate. In the rebuilt Portfolios the arithmetic itself stays on OddsShopper's one house method for now, so the choice you are actually making is about the inputs: whose prices count, and how much.
OS Weights: The Fair Line Out Of The Box
OS weights are the preset answer to the which-prices-define-fair question: OddsShopper's own weighting of the reference books that feed the fair line, maintained for you. In the rebuilt Portfolios, a portfolio is a named, saved filter state. Turn it on and its bets show up in your feed and it can notify you; leave it off and it quietly tracks its record in the background. The rebuild also puts the filters front and center on the main screen with quick-access tuning, shows your updated profit track record, and displays a daily swing range based on your portfolio settings.
[SCREENSHOT: Portfolios main screen — filters preset up front with quick-access portfolio tuning]
Consistency is how the preset earns its keep. When every bet in your portfolio is priced against the same reference-book weighting, your record answers a clean question: does this filter beat the market? Change the weights halfway through and you have quietly changed the ruler, so the first 200 bets and the next 200 are no longer measuring the same thing. For most bettors, most of the time, that consistency is worth more than any clever tweak, which is exactly why OS weights are the preset and not the compromise.
The launch is already drawing praise in public. @WSBTradesAlerts, a bettor who leans on the tool daily for player props, broke the rebuild down on X as a massive overhaul of PortfolioLab, the previous version of this screen, running through the updated track record, the custom-versus-preset devigging, the daily swing range, and the review tab. When someone who grinds a tool every day calls out the settings screen as a highlight, the settings screen is doing something right.
OddsShopper shops the line across 100+ sportsbooks and flags the bets priced better than the fair line, and the rebuilt Portfolios turn those flags into a tracked, devig-aware record. OddsShopper Pro unlocks the deeper tools with a 7-day free trial, and for new customers code CUSTOM30 takes 30% off your first payment once the trial ends.
Custom Weights: Building Your Own Fair Line
Custom weights are the craft of deciding, yourself, whose prices define fair. Flip the setting from OS weights to Custom weights and you choose which reference books feed your fair line and how much each one counts. In practice that means two big levers:
- Book Selection And Weighting. Instead of treating every book's price as equal evidence, you lean harder on the books that take real limits and move first, and discount the ones that copy lines and shade toward the public.
- Sharp-Book Anchoring. You weight your fair line around the sharpest market available, then treat everyone else's prices as candidates to beat rather than as evidence about the truth.
The difference is not academic. Picture a player prop where a sharp, limit-taking market like Pinnacle or Circa sits at -115/-105 while three slower recreational books post -110/-110. Weight every book equally and your fair line calls it close to a coin flip; anchor to the sharp book and your fair line tilts toward the -115 side, which changes which prices on the board count as +EV at all.
That is the rebuild's headline change: going custom means your fair line reflects your reference books instead of the house's, and every bet the portfolio flags is measured against the number you built. To be clear about what custom does not change: the vig-removal math itself stays on the house method for now — the setting reads "OS default," with more methods coming soon — so custom weights move the inputs, not the arithmetic.
[SCREENSHOT: fair-line settings — OS weights vs Custom weights, devig method on "OS default — more methods coming soon"]
Done well, that is a real edge. A bettor who lives in one corner of the market, say NBA player props or small-conference college football, can know things about which books are sharp in that niche that no one-size-fits-all method captures. Done casually, it is a slow leak, and the scoreboard problem makes it worse: the moment your custom line disagrees with the market, one of you is wrong, and your P&L will need hundreds of bets at minimum to tell you which.
Every setting you change is a claim about how the market works, and wrong claims produce fair lines that flatter bad bets. Custom devigging is not a shortcut to sharper bets; it is a responsibility, and it comes with homework we will get to in a minute.
When Custom Beats Preset (And When It Does Not)
The honest way to frame the choice is by what you actually know, not by how the options sound. Custom feels advanced and preset feels basic, and that instinct gets the decision backwards more often than it gets it right.
| Your Situation | Trust |
|---|---|
| You Bet Across Many Sports And Markets | Preset. One consistent method beats a patchwork of guesses. |
| You Are Still Building Your First Few Hundred Tracked Bets | Preset. You need a stable ruler before you can improve on it. |
| You Specialize In A Niche And Know Its Sharpest Books | Custom, graded against CLV every step of the way. |
| You Have A Specific, Testable Belief About Book Weighting | Custom, run beside a preset portfolio so you can compare records. |
| You Want To Tinker Because Tinkering Feels Sharp | Preset. Untested knobs are variance, not edge. |
Dwell on the second-to-last row for a moment, because it points at the move sophisticated users lean on: you do not have to choose once for your whole betting life. Portfolios are saved filter states, and a portfolio that is switched off still keeps score silently. Run your custom weights beside an OS-weights portfolio on the same markets, give them a real forward sample, and let the records argue instead of your instincts. One honesty note on that experiment: backtesting a custom weighting against history is a coming-soon feature, and when replayed records do exist they are labeled Replayed and kept separate from your Verified record — the tool never blends what you actually bet with what a simulation says you would have bet. That separation is the whole reason the record is worth trusting, and it turns the custom-versus-preset debate from a leap of faith into an experiment. Experiments need a grading metric, which is where the closing price comes in.
Sanity-Check Your Fair Line Against CLV
Closing line value is the gap between the price you took and the price when the market closed, and it is the fastest honest feedback a bettor gets. Sportsbook closing lines are the market's final, most informed answer, so consistently beating the close is strong evidence you are betting good numbers, long before your profit graph proves anything. If you have not internalized why, the CLV explainer is worth the detour.
CLV is also the referee for your devig choice. Remember the 4.8 points of vig you stripped out of that -110 market? Your fair line is only as good as what you stripped it toward, and CLV tells you whether the target was real. If your custom fair line is actually sharper than the preset, the bets it flags should beat the close at least as often. If your custom portfolio's CLV lags the preset's over a real sample, your model of the market is costing you money, no matter how sophisticated the settings feel.
The rebuilt review tab will not print you a CLV number, but it grades the ingredients of one, and it starts with more than your bottom line. Its headline is a verdict: it places your tracked result at a percentile of your portfolio's own outcome distribution — telling you, for instance, that a losing stretch sits at the 7th percentile of how this portfolio's bets tend to land, while flagging honestly that 17 bets is a small sample. Around that verdict, it breaks your record down three ways:
- Timing: when edges were posted versus when you actually placed your bets, including the EV you caught versus the EV you missed in that window.
- Price Capture: how much of the flagged edge your bet price actually kept by the time you clicked.
- Your-Picks Skew: whether your own pick selection pulls the record away from what the portfolio flagged.
WSBTradesAlerts summed the same idea up on X as the tab showing how much of your closing-line value is variance — something, in his words, that Pikkit doesn't even tell you. One reply in that thread made the same point: "short-term P&L is noisy, but consistently beating the closing price gives you much better EV." That is the mindset shift the review tab enforces: grade the process, not last week's results.
[SCREENSHOT: review tab — percentile verdict with the small-sample note, plus timing and price-capture breakdowns]
One caveat keeps you honest: CLV is evidence, not a verdict. A small sample of beating the close can be luck, and in slow markets the close itself can be beatable. Treat several hundred bets as the minimum before you declare your custom devig a success, and keep the preset portfolio running as the control the whole way.
The Fair Line You Can Defend
Back to the question we opened with: whose fair line should you trust? The answer the math supports is the one you can defend with a record. OS weights give you a consistent ruler from bet one, which is why most bettors should start there and stay there. The custom side is there for the moment you have a real thesis about which books are sharp, and the review tab exists so that thesis gets graded against the market instead of running on feel. Trust is not a setting you pick; a fair line earns it, a few hundred tracked bets at a time.
OddsShopper gives you the whole loop in one place: the live odds screen to shop the number, the +EV feed to find prices beating fair, and Portfolios to track whether your version of fair is beating the market. The 7-day OddsShopper Pro free trial is enough time to set up your first portfolio and watch the review tab go to work, and once the trial ends, code CUSTOM30 takes 30% off a new customer's first payment.
FAQ
What is a devig calculator?
A devig calculator removes the sportsbook's built-in margin (the vig) from a market's odds to reveal the no-vig, fair probability of each outcome. Comparing that fair line to the price you can actually bet is how you identify +EV opportunities.
What is a no-vig line?
A no-vig line, also called fair odds, is what a market's prices would be if the book took no margin. In a market with both sides at -110, the no-vig line on each side is +100, because each side is a fair 50% once the 4.8 points of vig are removed.
What is the difference between a custom devig and a preset devig?
The difference is whose prices define fair. The preset (OS weights) uses OddsShopper's own weighting of the reference books that feed the fair line, which keeps your fair lines consistent and your record comparable. Custom weights let you choose which books count and how much, so the fair lines reflect your judgment about where the market is sharpest instead of the house weighting.
What should I check before trusting a custom devig?
Check it against the closing price over a meaningful sample, ideally a few hundred tracked bets, and run it beside an OS-weights portfolio on the same markets as a control. If the custom fair line is real, its bets should beat the close at least as consistently as the preset's, and the review tab's price-capture and timing breakdowns should show you keeping the edge you were flagged, not leaking it.
