No secret sauce, just arithmetic

A good price only matters
when the game is worth playing.

The Deal Score is a sorting tool, not a review. It mixes what the current listing says with whatever game data we can get, and it is built so that no discount, however enormous, can rescue a game people did not like.

LootRadar scoring model

Five signals. One number.

0–100 Deal Score
Q
Quality35% of score
P
Price value25% of score
D
Discount20% of score
C
Review confidence10% of score
I
Player interest10% of score
Guardrail penaltiesThin evidence and unhappy players pull the total back down.
Final resultDeal Score

Quality leads on purpose. Discount strength hits a ceiling, so the sale badge runs out of road long before the game does.

Deal Score equals quality weighted at 35 percent, price value at 25 percent, discount at 20 percent, review confidence at 10 percent, and player interest at 10 percent, minus any guardrail penalties.

The five signals

What actually goes into the number

35%

Quality

Mostly Steam user sentiment, with critic scores chipping in when they exist. Either source can carry the signal alone, but a gap in the data never gets dressed up as certainty.

25%

Price value

Where we have it, the current price gets measured against a recorded historical low. In the list view a source-provided deal rating stands in, and it is labeled as the stand-in that it is.

20%

Discount strength

The discount contribution maxes out at 70% off. Go louder than that and you earn exactly nothing extra, which is the point.

10%

Review confidence

Review counts are scaled logarithmically, because going from 10 reviews to 1,000 tells you enormously more than going from 100,000 to 110,000.

10%

Player interest

Review volume plus a small nudge for recent releases. Popularity can back up a recommendation. It cannot stand in for the game being good.

Default quality gate

What gets through the door

To reach the main feed a game normally needs at least 70% positive reviews and at least 100 of them, assuming those numbers exist, plus either a real discount or a strong price-value signal. No critic score? Fine. That alone will not get a game hidden.

You can widen the review, price, and content thresholds whenever you like, and some curated collections run stricter rules than the defaults. Every filter choice gets written into the URL, so the exact view you built is bookmarkable and shareable.

Shown the door by default

  • DLC and season passes
  • Demos, prologues, and playtests
  • Soundtracks and art books
  • Virtual currency and cosmetic packs
  • Bundles and duplicate listings
  • Spam-like titles and suspicious list prices
  • Lower-confidence Early Access listings

Guardrails

Every penalty, in the open

No reliable quality data−18
Fewer than 10 reviews−22
10–49 reviews−12
50–99 reviews−5
Below 50% positive−25
50–69% positive−12
Non-game content−60
Spam-like title−15
Suspicious list-price ratio−10

Worked examples

Three deals, three very different fates

91Great game + big cut

94% positive · 48,000 reviews · 60% off

Quality 92.9, price value 92, discount 85.7, confidence 99.6, interest 86.4, penalties none. Everything pulls in the same direction. This is the shape of a deal you should probably just buy.

56Meh game + enormous cut

58% positive · 220 reviews · 90% off

The discount signal pins at 100 and it still does not matter. Quality lands at 56.3, mixed sentiment takes another 12 points off the top, and the whole thing misses the default feed. This is the exact case the score exists for.

82Beloved game + small cut

97% positive · 180,000 reviews · 25% off

Only 35.7 on discount, and it lands at 82 anyway. When a game is adored by that many people, a modest price cut is still a good day.

Sources & limits

What the score knows, and what it definitely does not

Current store and price listings arrive from a third-party pricing feed. Steam and SteamSpy metadata fill in review counts, popularity, and game details where they can. LootRadar layers its own eligibility rules and Deal Score on top of all that.

The published snapshot refreshes every three hours. Prices, availability, and metadata all drift in between, and the retailer’s checkout page decides what you actually pay.

Here is the honest gap list: no licensed full price history, no reliable record of how often a game gets bundled, no console pricing, no publisher release-pattern analysis, no verified Steam Deck status. When information is missing the score gets less confident. It does not get filled in with a guess.

LootRadar has no affiliate deals with retailers. Outbound links route through the pricing provider, which may earn a commission from the store. Either way, nobody can buy a change to the score, to what qualifies, or to the order things appear in.

Want to see what all that math actually surfaced?

Show me today’s deals