Jordan: "Make a slider for all transports." The all-armies run (reports/tiers/2026-10-03-all-armies.md) found dedicated transports ranked about 21% of the list lower than the reviewers put them, the Chaos Rhino 47 to 77% lower in all four Chaos armies and the Venom and Raider 85% and 77% lower: nothing in the ledger valued carrying a unit forward. This is one slider on the points ledger page, swept on every army with nothing else moved. Command: node tools/ledgercarry.js --w 0,0.25,0.5,1,2,4,8 --md (10 s). Our own numbers; unit names only, no rules text.
What was built
- The data.
tools/carry.jslists, for every army's ledger, each unit with the TRANSPORT keyword (the frozen build/units.json, checked against tools/ledger-inputs.lock.json) and the models it can carry: BSData's transport capacity, read by tools/transports.js (T-538) from the BSData checkout at the ledger's commit (374f505), the reading in the unit's own army's catalogue first (the Chaos Land Raider carries 14 in some armies, 12 in others). 73 transports in 14 armies; Blood Angels and Space Wolves have none in their pools. Numbers only, in tools/carry.json.tools/ledger-merge.jscopies it onto each form ascarry(the transport's capacity; also on the joined forms a transport is in, the Agents of the Imperium's Inquisitor riding with one), for all 16 armies. Every other byte of the ledgers is as it was: strippingcarryand its two meta lines gives back each file exactly. - The line. Carry = the models it can carry per 100 of its points ÷ the typical transport's (the median over the army's transports only, as Spawn's scale is over the units that spawn), × the Transports slider (
wCarry) × the Kill weight, × the army-wide rules like every line. So at 1, a typical transport's carrying is worth what a typical unit's Kill is worth.- Why in Kill's units: at the step-6 baseline Kill's weight is 498 and every other line's 1 to 19, so Kill is about 90% of a typical Value. A weight of 1 on the Carry line's own scale would add 1 point to a Value near 1,000: the sweep would move nothing.
- Capacity alone, not capacity × the typical passenger's worth: who would ride needs the transport's keyword rules against every unit of the army and a choice of passenger (the best, the typical, the cheapest), each a judgement the slider would hide; capacity per points already marks a big, expensive transport down (a Stompa carries 3 models per 100 points, a Trukk 20). It counts the space, not who rides. Not clearly better, so not done.
- A "Your call", hibernated: never fitted or searched, default 0. A form with no
carry, or the slider at 0, gets no Carry contribution at all, so the page is the step-6 baseline to the bit (tests/unit/ledgerpage.test.js's fingerprint). It stays out of the threat sum (the baseline's threat weighting is 0 anyway), so only transports' Values move: every change in pairwise below is a pair with a transport in it.
The sweep
The step-6 baseline (tools/ledger-benchmarks/2026-10-03-step6-baseline.json) with only wCarry moved. Pairwise: of the pairs a reviewer put in different tiers (S to D), the share the ledger orders the same way, the mean over the army's lists (the Tyranids' five and the Space Marines' two headline lists; every panel of the other 14), and overall the mean of the 16 armies' figures. Where the transports land: each graded transport's rank among the graded units as a share of them, from the top (10% is near the top), ours and the reviewers' (their mean letter over the army's lists, as the all-armies run's gaps); gap +: we place it lower than the reviewers. 60 graded transports, pooled; 26 of them dedicated.
| wCarry | 0 | 0.25 | 0.5 | 1 | 2 | 4 | 8 |
|---|---|---|---|---|---|---|---|
| Overall pairwise (16 armies) | 57.7% | 57.2% | 57.3% | 57.0% | 56.4% | 56.1% | 55.8% |
| 2+ tiers apart | 61.5% | 60.7% | 60.9% | 60.6% | 59.1% | 58.7% | 58.4% |
| ρ | 0.177 | 0.166 | 0.169 | 0.165 | 0.151 | 0.143 | 0.138 |
| Transports (60 graded): ours / theirs | 70% / 56% | 59% / 56% | 46% / 56% | 30% / 56% | 17% / 56% | 10% / 56% | 7% / 56% |
| their gap (+ lower than the reviewers) | +14% | +2% | −10% | −26% | −40% | −46% | −49% |
| mean distance from the reviewers' place | 27% | 29% | 29% | 34% | 41% | 47% | 49% |
| Dedicated (26): gap | +15% | +0% | −18% | −38% | −49% | −53% | −54% |
Per army: mean pairwise; the graded transports' mean gap.
| Army (lists, graded transports) | wCarry 0 | wCarry 0.25 | wCarry 0.5 | wCarry 1 | wCarry 2 | wCarry 4 | wCarry 8 |
|---|---|---|---|---|---|---|---|
| Tyranids (5, 3) | 74%; −1% | 72%; −26% | 72%; −34% | 70%; −53% | 68%; −62% | 67%; −72% | 66%; −83% |
| Space Marines (2, 11) | 59%; +12% | 59%; −1% | 59%; −16% | 58%; −31% | 57%; −43% | 56%; −46% | 56%; −46% |
| Adeptus Mechanicus (1, 2) | 69%; +17% | 69%; −2% | 68%; −11% | 65%; −30% | 63%; −47% | 61%; −59% | 60%; −61% |
| Agents of the Imperium (1, 3) | 45%; ±0% | 44%; −4% | 43%; −15% | 41%; −24% | 39%; −37% | 40%; −50% | 39%; −56% |
| Astra Militarum (1, 8) | 66%; +24% | 67%; +17% | 66%; +6% | 64%; −7% | 63%; −26% | 62%; −37% | 63%; −40% |
| Blood Angels (1, 0) | 54%; – | 54%; – | 54%; – | 54%; – | 54%; – | 54%; – | 54%; – |
| Dark Angels (1, 10) | 54%; +10% | 54%; −5% | 53%; −18% | 53%; −35% | 54%; −42% | 54%; −45% | 55%; −46% |
| Death Guard (1, 2) | 47%; +23% | 47%; +13% | 48%; −1% | 47%; −26% | 47%; −45% | 47%; −51% | 47%; −51% |
| Drukhari (1, 2) | 41%; +81% | 41%; +75% | 45%; +54% | 52%; +19% | 57%; −6% | 57%; −8% | 57%; −8% |
| Emperor's Children (1, 2) | 57%; +51% | 57%; +51% | 58%; +40% | 62%; +1% | 61%; −23% | 59%; −38% | 60%; −40% |
| Genestealer Cults (2, 2) | 69%; −22% | 68%; −28% | 68%; −30% | 69%; −32% | 69%; −39% | 68%; −45% | 68%; −53% |
| Leagues of Votann (2, 3) | 58%; +13% | 57%; +4% | 58%; −4% | 57%; −18% | 57%; −31% | 59%; −46% | 59%; −46% |
| Orks (1, 8) | 70%; −8% | 66%; −21% | 62%; −31% | 57%; −42% | 52%; −52% | 51%; −55% | 50%; −57% |
| Space Wolves (1, 0) | 54%; – | 54%; – | 54%; – | 54%; – | 54%; – | 54%; – | 54%; – |
| Thousand Sons (1, 2) | 46%; +37% | 45%; +22% | 48%; −1% | 49%; −17% | 49%; −33% | 49%; −37% | 49%; −42% |
| World Eaters (3, 2) | 62%; +33% | 63%; +24% | 62%; +14% | 61%; −7% | 59%; −36% | 58%; −54% | 57%; −57% |
What it says
- 0.25 puts the transports, as a group, where the reviewers put them (pooled gap +14% → +2%; the dedicated ones +15% → 0%). It costs half a point of overall pairwise (57.7% → 57.2%) and 0.011 of ρ: no setting gains overall, and none above 1 is close.
- One slider can't place them all, because the armies disagree about transports in opposite directions. The Chaos and Drukhari transports want 0.5 to 2: Drukhari 41% → 52% pairwise at 1 and 57% at 2, Emperor's Children 57% → 62% at 1, Thousand Sons 46% → 49%, Death Guard level at 0.5; at those settings the Chaos Rhino and the Venom come up to their letters. But in other armies the transports already sit at or above the reviewers' place at 0: the Space Marines' Impulsor and Rhino (2% and 24% of the list, the reviewers 34% and 83%), the Orks' walkers (Hunta Rig, Gorkanaut, Morkanaut), Genestealer Cults' Rockgrinder; Carry lifts them further and Orks lose 4 points at 0.25 and 13 at 1.
- The non-dedicated transports are most of the cost. Every TRANSPORT unit counts, as asked, so the Hierophant and Harridan (they carry Gargoyles; the reviewers grade them low) climb: the Tyranids' headline lists lose 2.2 points at 0.25 and 5 at 1. Measured on the side (not built): with Carry on dedicated transports only, overall pairwise is 57.5% at 0.25, 57.8% at 0.5 and 58.1% at 1, against 57.7% at 0, with the dedicated gap −4% at 0.5. A small gain, so a possible follow-up: one keyword check, still one slider.
- The mean distance from the reviewers' place grows from 27% at every setting: the slider moves every transport the same way, so it centres the group without tightening it.
The suggestion
Start Jordan at 0.25: the transports as a group land where the reviewers put them, for half a point of pairwise overall. Go to 0.5 to 1 to look at the Chaos and Drukhari lists, where the motivating misses close (Drukhari +11 points at 1), at the Orks' and Tyranids' cost. The default stays 0.
Each transport
Ours: rank among the army's graded units as a share of them, from the top; theirs: the same for the reviewers' mean letter.
| Transport (army) | Theirs | Ours at 0 | Ours at 0.25 | Ours at 0.5 | Ours at 1 | Ours at 2 | Ours at 4 | Ours at 8 |
|---|---|---|---|---|---|---|---|---|
| Hierophant (not dedicated) (Tyranids) | 88% | 81% | 33% | 21% | 2% | 2% | 2% | 2% |
| Harridan (not dedicated) (Tyranids) | 85% | 85% | 73% | 67% | 44% | 31% | 17% | 4% |
| Tyrannocyte (Tyranids) | 88% | 92% | 77% | 69% | 54% | 40% | 25% | 6% |
| Impulsor (Space Marines) | 34% | 2% | 1% | 1% | 1% | 1% | 1% | 1% |
| Rhino (Space Marines) | 83% | 24% | 2% | 2% | 2% | 2% | 2% | 2% |
| Repulsor (not dedicated) (Space Marines) | 19% | 58% | 37% | 15% | 4% | 4% | 4% | 4% |
| Land Raider Redeemer (not dedicated) (Space Marines) | 10% | 62% | 48% | 24% | 7% | 5% | 5% | 5% |
| Repulsor Executioner (not dedicated) (Space Marines) | 19% | 65% | 52% | 35% | 8% | 6% | 6% | 6% |
| Land Raider (not dedicated) (Space Marines) | 53% | 68% | 54% | 38% | 15% | 7% | 7% | 7% |
| Razorback (Space Marines) | 83% | 74% | 60% | 39% | 18% | 10% | 8% | 8% |
| Land Raider Crusader (not dedicated) (Space Marines) | 83% | 81% | 61% | 42% | 21% | 11% | 10% | 10% |
| Thunderhawk Gunship (not dedicated) (Space Marines) | 98% | 93% | 87% | 54% | 40% | 20% | 11% | 11% |
| Stormraven Gunship (not dedicated) (Space Marines) | 53% | 94% | 88% | 79% | 58% | 21% | 13% | 12% |
| Drop Pod (Space Marines) | 53% | 100% | 92% | 85% | 65% | 27% | 14% | 13% |
| Skorpius Dunerider (Adeptus Mechanicus) | 45% | 66% | 34% | 25% | 6% | 3% | 3% | 3% |
| Archaeopter Transvector (not dedicated) (Adeptus Mechanicus) | 86% | 100% | 94% | 84% | 66% | 34% | 9% | 6% |
| Sisters of Battle Immolator (Agents of the Imperium) | 65% | 58% | 54% | 23% | 8% | 4% | 4% | 4% |
| Corvus Blackstar (not dedicated) (Agents of the Imperium) | 65% | 65% | 62% | 58% | 50% | 27% | 15% | 8% |
| Imperial Rhino (Agents of the Imperium) | 65% | 73% | 69% | 69% | 65% | 54% | 27% | 15% |
| Taurox Prime (Astra Militarum) | 56% | 41% | 33% | 20% | 10% | 2% | 2% | 2% |
| Banehammer (not dedicated) (Astra Militarum) | 22% | 56% | 46% | 41% | 31% | 11% | 3% | 3% |
| Doomhammer (not dedicated) (Astra Militarum) | 22% | 59% | 57% | 46% | 36% | 13% | 5% | 5% |
| Stormlord (not dedicated) (Astra Militarum) | 22% | 61% | 59% | 48% | 38% | 15% | 7% | 7% |
| Chimera (Astra Militarum) | 56% | 75% | 62% | 52% | 39% | 16% | 11% | 8% |
| Taurox (Astra Militarum) | 56% | 90% | 75% | 62% | 48% | 26% | 18% | 10% |
| Valkyrie (not dedicated) (Astra Militarum) | 92% | 93% | 90% | 77% | 49% | 34% | 20% | 11% |
| Centaur RSV (Astra Militarum) | 56% | 97% | 93% | 87% | 74% | 52% | 21% | 13% |
| Impulsor (Dark Angels) | 18% | 9% | 4% | 2% | 1% | 1% | 1% | 1% |
| Repulsor Executioner (not dedicated) (Dark Angels) | 18% | 46% | 6% | 5% | 3% | 2% | 2% | 2% |
| Rhino (Dark Angels) | 92% | 49% | 38% | 28% | 6% | 3% | 3% | 3% |
| Land Raider Redeemer (not dedicated) (Dark Angels) | 18% | 51% | 43% | 31% | 10% | 8% | 4% | 4% |
| Land Raider (not dedicated) (Dark Angels) | 38% | 55% | 44% | 34% | 12% | 9% | 5% | 5% |
| Repulsor (not dedicated) (Dark Angels) | 56% | 67% | 46% | 35% | 14% | 10% | 8% | 6% |
| Razorback (Dark Angels) | 92% | 70% | 52% | 40% | 17% | 11% | 9% | 8% |
| Land Raider Crusader (not dedicated) (Dark Angels) | 75% | 81% | 62% | 42% | 19% | 13% | 10% | 9% |
| Stormraven Gunship (not dedicated) (Dark Angels) | 18% | 89% | 84% | 45% | 31% | 17% | 13% | 10% |
| Drop Pod (Dark Angels) | 92% | 100% | 90% | 73% | 53% | 22% | 14% | 12% |
| Chaos Land Raider (not dedicated) (Death Guard) | 71% | 69% | 64% | 47% | 14% | 3% | 3% | 3% |
| Chaos Rhino (Death Guard) | 39% | 86% | 72% | 61% | 44% | 17% | 6% | 6% |
| Raider (Drukhari) | 15% | 92% | 83% | 46% | 17% | 4% | 4% | 4% |
| Venom (Drukhari) | 15% | 100% | 96% | 92% | 50% | 13% | 8% | 8% |
| Chaos Land Raider (not dedicated) (Emperor's Children) | 63% | 96% | 96% | 78% | 22% | 9% | 4% | 4% |
| Chaos Rhino (Emperor's Children) | 30% | 100% | 100% | 96% | 74% | 39% | 13% | 9% |
| Goliath Rockgrinder (not dedicated) (Genestealer Cults) | 85% | 38% | 33% | 33% | 33% | 21% | 13% | 4% |
| Goliath Truck (Genestealer Cults) | 50% | 54% | 46% | 42% | 38% | 38% | 33% | 25% |
| Hekaton Land Fortress (not dedicated) (Leagues of Votann) | 8% | 40% | 40% | 35% | 30% | 20% | 5% | 5% |
| Sagitaur (Leagues of Votann) | 90% | 65% | 45% | 45% | 35% | 25% | 10% | 10% |
| Kapricus Carrier (Leagues of Votann) | 70% | 100% | 95% | 75% | 50% | 30% | 15% | 15% |
| Hunta Rig (not dedicated) (Orks) | 64% | 22% | 9% | 9% | 2% | 2% | 2% | 2% |
| Gorkanaut (not dedicated) (Orks) | 64% | 33% | 27% | 24% | 11% | 4% | 4% | 4% |
| Morkanaut (not dedicated) (Orks) | 64% | 42% | 36% | 27% | 13% | 7% | 7% | 7% |
| Stompa (not dedicated) (Orks) | 22% | 44% | 38% | 31% | 22% | 16% | 9% | 9% |
| Gunwagon (not dedicated) (Orks) | 64% | 64% | 47% | 33% | 31% | 18% | 11% | 11% |
| Trukk (Orks) | 96% | 84% | 51% | 38% | 33% | 20% | 20% | 13% |
| Battlewagon (not dedicated) (Orks) | 64% | 89% | 71% | 40% | 36% | 24% | 22% | 16% |
| Big'ed Bossbunka (not dedicated) (Orks) | 100% | 100% | 96% | 91% | 56% | 29% | 24% | 18% |
| Chaos Land Raider (not dedicated) (Thousand Sons) | 79% | 75% | 57% | 43% | 18% | 7% | 4% | 4% |
| Chaos Rhino (Thousand Sons) | 16% | 93% | 82% | 50% | 43% | 21% | 18% | 7% |
| Chaos Land Raider (not dedicated) (World Eaters) | 73% | 90% | 83% | 70% | 40% | 10% | 3% | 3% |
| Chaos Rhino (World Eaters) | 52% | 100% | 90% | 83% | 70% | 43% | 13% | 7% |
Decided alone
- Kill's units for the weight (above), so 0 to 8 is a useful range; the brief's "one typical unit's line" read as the line that carries a typical unit's Value.
- Every TRANSPORT unit, as Jordan said; a transport with several capacity readings takes its own catalogue's.
- Carry is outside the threat sum and multiplied by the army-wide rules like every other line.
- The 14 armies' compact summaries (docs/lab/progress/armies/) are unchanged: rebuilt with
--summary-onlyat wCarry 0 they come out the same, so they don't carry the field. Their full ledgers in build/ledger/ carrycarry.