Sources and method

What Chase 162 is

Chase 162 presents historical baseball season rosters and lineups, with sourced public data behind every player. In Classic Chase and the Daily Chase, you draft players from historical team seasons, fill a roster, and play a fictional 162-game season with those twelve picks. The archive lets you inspect the seasons behind those choices and follow their source links. It is independent and unofficial and is not affiliated with any league or club.

What is in the playable data

The current archive contains 2,554 team-season pages, spanning 1901 through 2025, with 103,108 roster entries in total. A roster entry counts a player on a particular team in a particular year; it is not a count of different people. The playable Chase catalog contains 2,541 draft seasons, spanning 1901 through 2025. An archive roster can exist without producing any draft candidates, so the archive and playable totals differ.

Across the playable season payloads there are 36,954 player-season candidates and 8,660 distinct player identities. Identity means the Wikipedia article title used by the extractor, rather than an independently audited registry of people. Returning players contribute candidates in multiple seasons, and a player appearing for multiple teams can contribute multiple candidates in a year. The draft prevents selecting the same article identity twice on your team.

The playable data has 31 distinct franchise or team codes. This counts the code groups used by the site, including historical names; it does not establish a definitive genealogy of franchise relocations. Hitters enter the draft pool only with at least 100 games in the source batting row and a matching roster player. Pitchers need at least 20 games and either at least 10 starts or 50 pitching outs, plus a roster match. These filters favor regular contributors and leave many real roster members outside the draft.

There are 12 roster slots: C (Catcher), 1B (First base), 2B (Second base), 3B (Third base), SS (Shortstop), LF (Left field), CF (Center field), RF (Right field), DH (Designated hitter), SP1 (Starting pitcher 1), SP2 (Starting pitcher 2), CL (Closer). Fielding eligibility comes from the position data joined by normalized name. There are 200 candidates flagged with unknown fielding positions; these hitters can fill any field position in Chase. That flexibility is a game rule, not evidence that they played every position. There are also 277 candidates flagged as approximate. These counts describe the generated payloads and are recalculated when the site builds.

How a rating is derived

Ratings compare a player's available statistics with the site's observed pool from the same year. The baseline collects batting and pitching rows across that year's available teams, identifying rows by team code and player article. Batting-average comparisons require valid hits and at least 50 at-bats. Pitching comparisons require ERA and at least 60 outs, equivalent to 20 innings. These rating thresholds are separate from the draft entry rules described above.

A baseline is marked observed only when it has at least 8 teams, 100 qualifying batters and 30 qualifying pitchers, with nonzero variation in batting average and ERA. Otherwise it is marked approximate. This is a coverage check within our extracted data, not a claim that the complete league has been recovered. Missing tables, uneven article detail and repeated team stints can affect the comparison pool. Years with richer source tables can support more comparisons than years with sparse articles.

For a hitter, the score combines the player's distance from the same-year mean batting average with the distance from the mean home-run rate per at-bat. Batting average receives 80 percent of the weight and home-run rate receives 20 percent. The result is pulled toward the middle according to at-bats divided by at-bats plus 150. A pitcher uses the distance below the same-year mean ERA, pulled toward the middle according to innings divided by innings plus 90. Lower ERA increases that score.

Both source formulas start at 50, apply a scale of 12 to the adjusted comparison, round to an integer and limit the output to 1 through 99. Archive ratings use that scale; Chase OVR applies the catalog percentile curve described below. A missing home-run input contributes no home-run advantage to the hitter score. A missing required row is unrated, while inadequate required statistics produce an insufficient-sample status. The Chase extractor can still admit an otherwise eligible candidate with a neutral source rating of 50 and mark it approximate before applying the display curve.

The simulator uses derived batting probabilities and a pitching factor, rather than simply comparing OVR badges. It shrinks and limits those inputs to keep the model playable. Missing walk, double or triple inputs can use baseline or formula defaults; an eligible hitter without a usable rating gets neutral batting probabilities, and an unrated pitcher gets a neutral factor. These are explicit modeling assumptions. They do not fill in missing source statistics. The Chase extractor also treats missing games and saves as zero for selection and relief checks; its saves summary can show that default. It should not be read as a verified historical zero. The formula emphasizes contact, power and ERA; it does not measure every skill that makes a real player valuable.

How the rarity label works

A selection can carry a rarity label, Elite or Rare. The comparison uses source stat rows from the same year, without requiring a matching roster player. Hitter rows need at least 100 games, at least 50 at-bats and positive runs batted in. Pitcher rows need at least 20 games, at least 50 numeric pitching outs and positive strikeouts. These comparison pools differ from the draft entry pool, and each pool needs at least 20 rows or its labels are suppressed. A hitter is Elite at or above the 99.95th-percentile RBI cutoff and Rare at or above the 98.5th; pitchers use the same cutoffs for strikeouts. Cutoffs include ties, so the bands need not contain exactly 0.05% or 1.5% of the rows. Missing counting statistics produce no label.

The label is deliberately not an award, and it describes a single season rather than a career, so a great player can be labelled in several seasons without the label becoming rarer each time. Only the two top bands are shown. A third, broader band was tried and left a label on about a third of the roster, which made an ordinary player look noteworthy and buried the ones worth picking; most players therefore carry no label at all.

How the shared report card is derived

A shared season page shows an overall score out of 100, six letter grades and seven profile bars. The draft screen's TEAM METRICS panel shows OVR, OFF, DEF and PIT as the required positions are filled. OVR averages selected player ratings, OFF uses the same position-weighted offense as the card, DEF remains field fit (70 for a known position, 40 for an unknown one), and PIT averages the two starters and closer.

Offense is the rounded weighted mean overall rating of the nine hitters. Position weights are C: 1.6, 1B: 0.9, 2B: 1, 3B: 1.1, SS: 1.6, LF: 1, CF: 1.1, RF: 1, DH: 0.5. Catcher and shortstop carry the most weight because they are harder to replace; the designated hitter carries the least. Rotation averages the two starters and Bullpen uses the closer's rating. Card Defense averages the eight fielders' ratings using the same position weights, excluding DH; an unknown position retains the 40-point fallback. This estimates quality from overall ratings, not measured defensive runs. Weakest shows the lowest-rated of all twelve picks, with name, slot and grade; ties follow roster slot order.

Balance takes the lower of one hundred minus the spread between offense, rotation and bullpen, or the lowest of those three lines minus half that spread rounded to an integer, then floors the result at zero. The stricter reading is used because spread alone would score a uniformly weak roster as perfectly balanced. Balance appears on the report card, not in TEAM METRICS. Letters follow fixed bands: A+ at 90 and above, A at 85, B+ at 80, B at 75, B− at 70, C+ at 65, C at 60, C− at 55, D+ at 50, and D below that.

Chase OVR maps each candidate's source rating to its empirical percentile across the entire draft catalog. Ties share their middle rank. Percentile-to-OVR anchors are 0%: 30, 5%: 40, 25%: 50, 50%: 60, 75%: 70, 90%: 80, 95%: 85, 99%: 95, and 100%: 100. Scores interpolate between anchors and round to integers. These are relative ratings within the extracted catalog, not win probabilities or new historical statistics. The simulator retains the original source rating for its existing draft-form boost, so changing the displayed curve preserves seeded game scores.

The overall score is a fixed weighted mean: offense ×1, rotation ×2, bullpen ×2, defense ×1.5, balance ×0.5, dominance ×0.5, weakest ×1.5, and record ×27. The weights sum to 36; record contributes 75% and roster quality contributes 25%. Weakest contributes 4.17% overall, funded by reducing Balance and Dominance, so one weak pick matters without controlling the team grade. Dominance is the rounded mean of all twelve ratings plus one and a half times their spread, rounded and capped at 100, so a top-heavy roster scores above a flat one. The record score interpolates linearly between fixed wins-to-score anchors: 0 wins: 0, 81: 35, 95: 55, 120: 65, 152: 82, and 162: 100. Only a perfect record reaches 100. These anchors calibrate the headline to the game's outcome bands: roughly 158+ wins for A grades, 140–152 for B/B+, 120–138 for C+/B−, 95–118 for C−/C, and below 90 for D grades. Roster quality can move a team near a boundary into an adjacent band. The weighted sum is divided by 36 and rounded once, to one decimal place.

The card names a Standout player as the highest rated of the twelve, and a Most balanced pick as the selection standing behind whichever of offense, rotation or bullpen came out lowest. The site does not compute wins above replacement, so these callouts describe the site's own rating rather than claiming a measure this project does not calculate. The biggest weakness names the lowest of Offense, Rotation, Bullpen and Defense. Balance is derived from other lines, and the separate Weakest row identifies an individual player; either can score below this team-line callout.

How the season simulation works

A Chase regular season has 162 games. The seed controls the shuffled draft seasons, opponent selection and game randomness. Given the same seed, data version, picks and weather rules, the same scores are reproduced. The seed alone does not make different rosters equivalent. Saved game links carry the choices and data version; a link from an incompatible version is rejected instead of silently substituting different players.

The schedule selects up to 18 historical opponents that can field a team and cycles through them. Home and away alternate, as do SP1 and SP2. Each game uses a single pitching factor combining the chosen starter with the closer, weighted as six starter innings and three relief innings. Those weights are a simplification; the engine does not execute a bullpen manager's decisions or track fatigue between games.

Within a game, seeded draws produce outs, walks, singles, doubles, triples and home runs from the batter's probabilities adjusted by the opposing pitching factor. Runners advance by simple fixed rules. Games can continue beyond nine innings, with a safety limit of 30 innings; a remaining tie is resolved by a seeded extra run. That tiebreak is a game mechanism, not a reconstructed historical play. The engine does not model stolen bases, fielding errors, injuries or tactical substitutions.

Each draft starts with three team respins and three era respins. A team respin keeps the source year and changes the franchise; an era respin keeps the franchise and changes the decade. Respins replace only the current offer. Every drafted player keeps their identity and position, and later initial draws keep their original sequence. A token is spent only when a different eligible offer is available. Your twelve drafted players stay on the roster for the entire season. Once your roster is complete, play all 162 games and see the result. First-half and second-half records summarize the season without requiring another decision.

Weather in Chase 162

New games draw Clear, Rain, Snow, Wind, Haze, Cold or Heat from the seed and game number. Wind also draws a tailwind or headwind. The player does not choose the weather, and a share link reproduces it. Legacy saved games keep their original scores with weather off.

These percentages are game-design tuning informed by documented baseball conditions, not measurements from a study or empirically derived effects. Both teams receive the same condition. Weather changes scoring only: it does not cancel, postpone or shorten games. Every regular season still has 162 games, and the goal is still a perfect 162.

Nominal weather modifiers at severity 1.0; effects scale per game from 0.4 to 1.6
ConditionBatting probabilitiesPitching strengthFatigue multiplier
Clear0%0%1.0
Rain-6%-3%1.0
Snow-9%-2%1.0
Wind+7% tailwind; -7% headwind0%1.0
Haze-4%-4%1.5
Cold-2%-8%1.0
Heat-3%-10%1.0

Each game draws a seeded severity between 0.4 and 1.6. Severity multiplies the table's batting and pitching effects; 1.0 gives the nominal values above. Batting modifiers scale hit and walk probabilities, preserving the mix of hit types. Weaker pitching increases the engine's factor for offense allowed. Fatigue multiplies the pitching penalty within the game: Haze therefore has a nominal -6% pitching-strength adjustment before severity. There is no fatigue carried between games. Wind's pitching adjustment is zero; heat uses the listed pitching penalty as a simplified fatigue effect.

Rain has affected official results: the 2020 Yankees-Nationals opener ended during the sixth inning with New York ahead 4-1. In the regulation-game framework, five completed innings make a game official; a home lead can make it official after the visiting team has made 15 outs. Later weather can still require suspension in some tied or incomplete-inning situations. Chase always plays the game through.

Snow is grounded in the April 9, 1996 Yankees-Royals home opener at Yankee Stadium, played in snow with temperatures in the low 30s; New York won 7-3. Jeter's first career home run was April 2 in Cleveland, a separate game.

Haze is informed by the September 14, 2020 Athletics-Mariners doubleheader in Seattle. Reported air quality was 242 before play and exceeded 280 during it; Oakland pitcher Jesus Luzardo described breathing difficulty. MLB had no stated AQI cutoff in that report. Cold is informed by the April 8, 2003 Yankees-Twins game, with a first-pitch temperature of 35 F and a wind chill around 28 F.

Fenway Park and the old Yankee Stadium are open-air parks where wind can affect play; the seeded direction is a simplified game rule. Heat's effect on pitching is likewise a design assumption. These historical examples establish that teams play in difficult conditions; they do not quantify the modifiers above.

What this is not

Ratings, OVR and simulated records are not official statistics and are not predictions. The project does not ingest official OPS+, ERA+ or WAR fields into its rating formulas. Same-year normalization is a limited comparison within the extracted data; it does not correct for every ballpark, league, opponent, rule change or historical context. A strong badge should be read as a result of the published formula, not an authoritative ranking across baseball history.

Missing source rows remain visibly missing. We do not interpolate historical hits, innings or other missing source statistics, and an unavailable source value remains unknown. The simulation defaults described above are separate from that record. A team may have roster data without usable statistics for every player, and eligibility may be uncertain when the position join fails. Staff are excluded from player choices, but parsing and name matching can still miss or misidentify information.

Draft DEF is a position-coverage proxy; report-card Defense estimates quality from overall player ratings. Neither measures defensive runs. The model omits much of the sport and has no published predictive validation score; its real-world forecasting accuracy is unknown. Results are useful for replaying choices under consistent game rules. They cannot establish what a historical team would actually have won with a different roster. If a row looks wrong, use its source article and retrieval timestamp to investigate, and report a data correction.

Data sources and licensing

Season rosters and available batting and pitching tables come from the Wikipedia season article linked on each season page. Each page shows its own retrieval timestamp. We exclude staff from player choices, join statistics to roster players by Wikipedia article title, convert baseball innings notation to outs, and derive approximate ratings from available same-year statistics. Missing source rows remain visibly missing; simulation results are fictional.

Wikipedia text is available under Creative Commons Attribution-ShareAlike 4.0. Each season page links directly to its source article and states the transformation applied.

Fielding eligibility uses a separate position snapshot built from Lahman Appearances and People tables, joined by year, team and normalized name. That input's license metadata is unknown in this repository; the Wikipedia license statement describes Wikipedia-derived material. A failed position join stays unknown rather than becoming a claimed historical appearance.

The displayed tables are transformed snapshots, so a source article can change after retrieval. Attribution points to that article rather than suggesting Wikipedia reviewed this game's ratings. When reusing source-derived material, retain the source attribution and follow the linked license terms. Source facts and the site's derived ratings should remain distinguishable in any reuse.

Machine-readable index

The machine-readable season index lists each built roster's team code, year, source title, historical team name, available stat tables, and player count. It is the archive index, not a claim that every listed roster can supply every draft position. Table-availability flags say a table exists; they do not certify that all rows or all fields are complete.

Use the season filter to locate individual rosters, or browse by decade for static archive links. Each season remains available with its attribution and source context. The index makes it possible to inspect coverage before drawing conclusions from the game, while the season pages provide the detail needed to check a particular player or year.