Our value list runs on a simple engine: when two independent live trackers agree on where an item sits, the number moves. Most weeks that engine just runs. Right now it is stuck in the most instructive way possible, and the stuck part is worth learning from.
Since mid-August, the weekly cross-check has had Lightning and Gas in a genuine deadlock. fantasyblox holds Gas above Lightning. tradekitsune holds Lightning above Gas. Both trackers are current, both updated this week, and they have held the same opposite sides at every check where sides were recorded. This is not the situation covered in why value lists disagree, where two sites quote different magnitudes for the same item and the gap measures uncertainty. And it is not the silence problem from when a tracker stops listing your item, where a source stops making claims at all. A split is both sources speaking clearly - and contradicting each other on order.
A split freezes one rung, not the ladder
The tempting reads are all wrong. You do not average the two positions, because the midpoint is a number neither source published - you would be inventing evidence and signing their names to it. You do not pick a side, because a 1-1 tie contains no information about which claim is better. And you do not freeze everything and wait, because the deadlock is about one comparison, not the whole list.
Watch what actually happened on August 13. Both trackers agreed Lightning belonged above Dough, Venom and Spirit, so Lightning moved up sharply - those rungs were unanimous. One tracker also put it above Gas; the other did not. So the agreed part of the move happened in full, and the disputed rung went on watch instead of into the number. Three weeks later Lightning still sits where the unanimous evidence put it, Gas still sits where its own evidence put it, and the one comparison the sources fight about is the one comparison our list refuses to rule on. A split is surgical. It stops exactly the claim that lacks agreement.
Why a third source can't break the tie
The obvious fix - find a tiebreaker - fails on inspection. The only other list we know of stopped updating months ago, which disqualifies it as evidence of anything current. But the deeper problem would remain even if a fresh third list existed: recruiting a judge after you know the deadlock is shopping for a verdict. If a third source had been in the weekly rotation all along, its vote would simply count. Adding one mid-dispute, because it happens to break the tie the way you would like, turns verification into decoration. The honest options are two: wait for one of your existing sources to change its mind, or live with the split.
Write the sides down
Here is the embarrassing lesson, straight from our own ledger. In the split's first week, we logged that a disagreement existed - but not which source held which side. When week two arrived and the sides were visible, there was no way to say whether the trackers had swapped positions or held them, and a swapped-sides claim had to be retracted. The network owned that failure publicly in Four weeks of not knowing, and the fix is now procedure: every check records the sides by name, with dates.
The reason this matters to you: stable sides and swapping sides are different markets. Sources that hold the same opposite positions week after week are reading genuinely different communities or methods - a standing disagreement that says the item's order is truly contested. Sources that trade sides are measuring noise. You can only tell these apart if you wrote down who said what, when.
Splits end - that's why the discipline pays
A deadlock is a state, not a verdict. The proof sits one rung up the same list: Control spent three weeks in a disputed ordering against Gas, and on August 13 both trackers came out placing Control above Gas - and above Tiger and Yeti, directly below Kitsune. The dispute resolved, the value moved the same day, and the move was trustworthy precisely because it had waited for agreement instead of splitting the difference weeks earlier.
If you hold an item your own two references disagree on: price off the neighbors they agree about, hold the disputed comparison, date your notes, and do not let a counterparty cherry-pick whichever list flatters their side - that trap and its answer are the heart of how to read a value list. One mechanical warning while you compare: some trackers' list views mix variant and permanent rows in ways that invert the base order, so read the per-item pages, not the list page, before concluding two sources disagree at all. Half the splits people report are really one misread list view.
Check where everything stands on the full value list - every number dated, every dispute noted, and one rung, as of this writing, honestly frozen.