BAMservatory emblem

BAMservatory

Solana MEV
BAMsey, the Observatory sentinel BAMsey · Alert · Nakamoto 3

An independent transparency & early-warning layer for Jito's Block Assembly Marketplace.

Sentinel: alert · Window 2026-06-20 → 2026-10-29 · 158,005 snapshots @ ~60s
Generated 2026-10-29T21:20Z · source: public BAM API

BAM at a glance

Live consensus probe
Solana stake routed
36.6%
of ALL Solana stake routed through BAM
159.9M SOL−0.94pp 7d
Connected validators
375
validators connected
Spreadacross 15 regions
Active infrastructure
17
BAM nodes live
Peak regionfra (95 vals)
Top node density
23.0%
stake on the top node
ams-mainnet-bam-2-teeAlert
BAMsey BAMsey's read Synthesis
Figures cross-checked

BAM share fell 0.01 percentage points over the last 23 hours and 0.92 percentage points over 7 days, leaving 36.6% of all Solana stake routed through BAM. Concentration remains elevated: ams-mainnet-bam-2-tee holds 22.4% of BAM stake, and the node Nakamoto coefficient is 3. Recent precursor signals are watch items, not evidence of another structural rollover.

Written by gpt-5.5-2026-04-23 from the published metrics · ✓ every figure cross-checked against the dataset before publishing 2026-10-29T16:39Z

Decentralization — how concentrated is BAM?

Entropy analysis
Nodes metricCritical
3
Nakamoto coefficient (nodes)
min nodes controlling >50% of BAM stake
Validator entropyNominal
26
Nakamoto coefficient (validators)
min validators controlling >50%
Geographic spreadCritical
3
Nakamoto coefficient (regions)
geographic concentration
Validator core top 10Index
28%
held by the top 10 validators
top 1: 8.1% · top 5: 19.5%
BAMsey in alert state
Concentration alert — node Nakamoto coefficient is 3.Threshold breached

Just 3 of the 17 live BAM nodes hold a majority of marketplace stake, the largest holding 23.0% on its own — and the top node by stake changed hands 68 times in this window, driven by a handful of whale validators toggling between nodes (see Whale Watch). These are exactly the concentration dynamics a transparency layer should surface.

Verification — checking what BAM reports

Independent oracle match

Everything above is gathered from BAM's own API, which makes it an index of what BAM says. This section checks it. Stake is verified against Solana itself, where the chain is the authority and BAM's figures either match or they do not. Membership is cross-checked against Jito's separate Kobe API, which publishes the same fact independently.

Read at 2026-10-29 21:20:14 UTC, on its own slower cycle than the figures above — it queries three services and a full Solana vote-account set, so it runs less often than the 60-second capture. Counts here will therefore differ slightly from the headline, which is the more recent reading, not a contradiction of it.
State cross-check
Matches
BAM's reported stake vs Solana
375 validators checked · typical deviation 0.0000% · worst 0.0000%
Deviation margin
0.3166pp
share gap, and it is the network total
BAM publishes 36.5666% · chain gives 36.8832% · the same 159,865,198 SOL divided by network totals 3,752,768 SOL apart
Kobe dual-source
Agree
Jito's two sources agree
BAM explorer lists 375 · Kobe flags 375
Dispute resolution
0
SOL under disagreement
stake attached to the validators in dispute
Validators the two sources disagree on
onlyExplorer: 0 – 282
2026-08-10Cur: 02026-10-29
Typical per-validator stake deviation (%)
stakeMedianRelPct: 0.0000% – 0.0000%
2026-08-11Cur: 0.0000%2026-10-29
What this does not establish

Which validators run BAM still comes from Jito — a BAM-produced block is indistinguishable from any other on chain, because BAM changes how a block is assembled, not what ends up in it. That membership claim is now cross-checked between two of Jito's own systems rather than taken on faith, and the stake attached to it is verified outright against Solana — but it is not independently derived, and no amount of cross-checking makes it so. That gap closes with published attestations, not with more sources. Re-checked automatically; still no attestation endpoint as of 2026-10-29.

Early warning — structural rollover detection

Predictive telemetry
22minlead time
FRA structural rollover ― fra-mainnet-bam-1-tee → fra-mainnet-bam-2-tee

Precursor fra-mainnet-bam-2-tee appeared at 05:36Z; cutover at 05:58Z on 2026-06-24. 9 regional precursor signals fired across the coordinated event.

BAM periodically migrates validators between TEE nodes in coordinated, region-by-region rollovers. The Observatory detects these before they complete: when a new node appears in a region, a cutover in that region typically follows within ~30 minutes. Validated on the 2026-06-24 event with 22-minute lead time (n=1 structural event; detector is live and accumulating more). Live "leadership flips" below are mostly whale-driven stake toggles, not structural rollovers — the Observatory labels them as such rather than counting them as early-warning wins.

Precursor signal log12 recent ingested events
Time (UTC)EventTypeDetail
10-29T16:27 SIGNAL precursornew ams-mainnet-bam-dev
10-29T16:02 SIGNAL_XREGION precursornew tyo-mainnet-bam-dev-tee
10-29T15:39 SIGNAL precursornew ams-mainnet-bam-dev
10-29T02:42 SIGNAL_XREGION precursornew tyo-mainnet-bam-dev-tee
10-28T15:30 SIGNAL_XREGION precursornew tyo-mainnet-bam-dev-tee
10-28T14:35 SIGNAL_XREGION precursornew tyo-mainnet-bam-dev-tee
10-28T06:29 SIGNAL_XREGION precursornew tyo-mainnet-bam-dev-tee
10-27T22:01 SIGNAL_XREGION precursornew lon-mainnet-bam-1-tee
10-27T18:23 SIGNAL_XREGION precursornew tyo-mainnet-bam-dev-tee
10-27T16:44 SIGNAL_XREGION precursornew tyo-mainnet-bam-dev-tee
10-26T13:49 SIGNAL precursornew ams-mainnet-bam-dev
10-24T17:08 SIGNAL_XREGION precursornew tyo-mainnet-bam-dev-tee

Current topology — 17 nodes

TEE fabric overview
NodeRegionValidatorsStakeShare
ams-mainnet-bam-2-tee ams 6736.79M 23.0%
fra-mainnet-bam-2-tee fra 9526.38M 16.5%
sqq-mainnet-bam-2-tee sqq 2619.56M 12.2%
tyo-mainnet-bam-2-tee tyo 4118.00M 11.3%
iad-mainnet-bam-1-tee iad 914.11M 8.8%
sin-mainnet-bam-1-tee sin 3513.78M 8.6%
sea-mainnet-bam-1-tee sea 25.80M 3.6%
ewr-mainnet-bam-1-tee ewr 215.38M 3.4%
pit-mainnet-bam-2-tee pit 153.99M 2.5%
dub-mainnet-bam-1-tee dub 43.85M 2.4%
lax-mainnet-bam-1-tee lax 133.15M 2.0%
hkg-mainnet-bam-1-tee hkg 173.10M 1.9%
lon-mainnet-bam-1-tee lon 142.57M 1.6%
slc-mainnet-bam-1-tee slc 91.32M 0.8%
dfw-mainnet-bam-2-tee dfw 51.22M 0.8%
tyo-mainnet-bam-dev-tee tyo 10.52M 0.3%
ams-mainnet-bam-dev ams 10.35M 0.2%

Whale watch — who controls BAM stake

Routing power
#ValidatorNode regionStakeShare
1DRpb…21hy iad12.88M 8.06%
25Cch…q9Ux sea5.72M 3.58%
3Fc6N…ULGD sin4.52M 2.83%
49jxg…nGFP fra4.48M 2.80%
55EhG…ULeC tyo3.60M 2.25%
6GzDf…4ZzG dub3.38M 2.12%
76y7V…syrK sqq2.92M 1.82%
8GBQ2…AAAA ams2.71M 1.69%
9GSTa…3XKc ams2.50M 1.57%
105ejb…Yj1z sqq2.50M 1.57%
11GQzM…gV46 tyo2.50M 1.57%
12BSVc…Edwq sqq2.50M 1.57%

Stake leadership of the BAM network is steered by a small set of large validators. Surfacing who they are and where they route makes BAM's power distribution legible to the Solana ecosystem — a public good no tool provides today.