Do the maths with us · Regional Court Frankfurt am Main · Case no. 2-10 O 2/26
Asymmetry Twelve court errors complained of. All in the same direction. Coincidence?
Mishaps happen, in courts too. But genuine mishaps are scattered: sometimes they hit one side, sometimes the other. In our case we complain of twelve errors by the court, and all twelve work to the detriment of the claimant. The calculator shows at once how rare it would be for all of it to come together by chance in a single case. It starts with values that treat the court far more generously than the sources do; you can switch to the source values with one click. You can change every value yourself.
The idea in 20 seconds
Chance spreads out. These twelve errors do not.
Toss a coin twelve times and it lands sometimes heads, sometimes tails, usually roughly half and half. Landing on one particular side twelve times in a row happens only once in 4,096 series. Our twelve complaints all fall on the same side.
Important: This page reflects the claimant’s view in pending proceedings. The twelve points are our complaints about the court’s conduct of the proceedings, not findings of the court; nothing has been decided. There are no official statistics on procedural errors: all values are disclosed, deliberately generous assumptions that you can change yourself. The result does not prove intent. It shows how poorly the explanation “nothing but independent coincidences” fits the court file.
The calculator
The result – and your own assessment
Each slider shows in what percentage of cases something like this happens. Every scale carries two markers: ▲ the conservative starting value, clearly in the court’s favour, and ▲ the standard value from the sources. Move the mouse over a marker to see how it is derived; a click sets the slider to that value. At the top you switch all sliders at once. The scale runs from 0.1 % to 100 %; at 100 % a point drops out of the calculation entirely.
Why do we not start with the source values? Because we do not need them: even with values that go a long way towards the court, the result remains rarer than tossing heads 18 times in a row.
All twelve errors by chance in the same case
p = 1.89 · 10⁻⁶as rare as tossing heads 19 times in a row
Less likely than tossing heads 18 times in a row (1 : 262,144). But more likely than being dealt a royal flush.
Without JavaScript the calculator shows the result with the conservative starting values.
- Service & court file
- Lawful judge
- Case management & time
Group · 2 events · one factor
Tracking histories: not requested, not in the court file
Two events, one topic: the tracking histories of the deliveries are neither requested nor added to the court file.
Service & court fileTracking histories stay outside the court file
Three times we asked the court registry to add the electronic tracking histories of the deliveries to the file. To this day they are not in the file.
Details, legal basis and evidence
On 09., 13. and 17.03.2026 we asked the registry to retrieve the electronic tracking histories and add them to the file. They would show where the statement of claim was for four weeks. The answer came on 18.03.2026 from the judge then in charge: tracking data would not be requested (see next point).
In our view, data generated in the course of service belong in a complete court file.
- Dates
- Requests 09., 13. and 17.03.2026
- Legal basis
- Complete record-keeping (§§ 182(3), 298a ZPO), inspection of the file (§ 299(1) ZPO)
- Evidence
- our letters of 09./13./17.03.2026 B01 B02, p. 1
- Effect
- to the detriment of the claimant: primary data on service remain inaccessible
Service & court fileFirst refusal: “Tracking data will not be requested”
On 18.03.2026 the judge initially in charge refused to retrieve the tracking histories. Only the certificates of service were said to be decisive.
Details, legal basis and evidence
“Only the certificates of service, which you have, serve as proof of service pursuant to § 182 ZPO. Tracking data – insofar as any exist at all – will therefore not be requested.”Court, order of 18.03.2026 (B01, p. 1) – translation; the German original prevails
But it is precisely the accuracy of these certificates that is in question. In our view, a court may not declare in advance that the evidence capable of checking them is irrelevant.
- Date
- 18.03.2026
- Legal basis
- Right to be heard (Art. 103(1) GG); rebuttal evidence against public documents (§ 418(2) ZPO)
- Evidence
- B01, p. 1
- Effect
- to the detriment of the claimant: rebuttal evidence is cut off
One factor for both events in this groupConservative starting value: 12 %In the court’s favour: three times the source value, at most 50 %. The calculator starts with this value.
Standard value from the sources: 4.0 %There are no official statistics for these errors. As a group they count once, with the highest individual value in the group: 4 out of 100 regional court cases.
Source: Claimant’s tolerance model; events documented by B01, B02
Service & court fileSecond refusal – with the charge of a “shot in the dark”
After the change of judge, the court maintained the refusal on 03.07.2026. Of our application it said that it was “likely … made as a shot in the dark”.
Conservative starting value: 9.0 %In the court’s favour: three times the source value, at most 50 %. The calculator starts with this value.
Standard value from the sources: 3.0 %There are no official statistics for this type of error. The source value is the claimant’s generous assumption in the court’s favour: 3 out of 100 regional court cases.
Source: Claimant’s tolerance model; event documented by B02
Details, legal basis and evidence
“The court file is not incomplete. The tracking data are therefore not to be added to the file.”Court, 03.07.2026 (B02, p. 1) – translation; the German original prevails
Of our application for a finding that the service process had been interfered with, the court says that it was “likely … made as a shot in the dark” (B02, p. 2; translation). Yet the data that could substantiate our submission are kept out of the file by the same court. We see this as an anticipatory assessment of evidence.
- Date
- 03.07.2026 (after a new judge took over the case)
- Legal basis
- Art. 103(1) GG; § 286 ZPO
- Evidence
- B02, pp. 1–2
- Effect
- to the detriment of the claimant: second block on the same source of evidence
Group · 2 events · one factor
Composition: single judge instead of chamber
Two events, one topic: neither at the outset nor after the change of judge is any consideration by the chamber of the questions of fundamental importance documented.
Lawful judgeSingle judge instead of chamber – from the outset
The proceedings have been conducted before a single judge from the start. Given their fundamental importance, we consider them a matter for the chamber.
Details, legal basis and evidence
On 17.02.2026 the hearing was scheduled “before the single judge”. Whether this rests on original competence (§ 348(1) ZPO) or on a transfer by the chamber (§ 348a ZPO) is not apparent from the file available to us. We requested the documents on 20.10.2026.
In our view, the case is of fundamental importance: six pillars of the claim and a sales organisation that, by its own account, serves around 8 million customers. Such cases belong before the chamber (§ 348(3) sentence 1 no. 2, § 348a(1) no. 2 ZPO). The composition of the court can, however, be challenged only to a limited extent (§ 348(4) ZPO).
- Date
- Hearing scheduled 17.02.2026
- Legal basis
- § 348(3) sentence 1 no. 2, § 348a(1) no. 2 ZPO
- Evidence
- B03 B16, p. 4 K7.42 B33, pp. 80–81
- Status
- Basis of allocation not documented, requested on 20.10.2026
- Effect
- to the detriment of the claimant: no deliberation by the chamber
Lawful judgeNew single judge – no documented referral to the chamber
Since 29.04.2026 a different single judge has been deciding. We are not aware that the chamber has dealt with our questions of fundamental importance.
Details, legal basis and evidence
On 23.03. and 27.04.2026 we set out in detail why the case is of fundamental importance. Since 29.04.2026 a different single judge has been deciding. We are not aware of any documented consideration of these questions by the chamber (§ 348(3) ZPO).
Whether any review took place is not established. We requested the relevant documents on 20.10.2026.
- Date
- from 29.04.2026
- Legal basis
- § 348(3) ZPO
- Evidence
- B04 B02, p. 1 B10 B33 B16, p. 4
- Status
- Documentation requested on 20.10.2026
- Effect
- to the detriment of the claimant: questions of principle stay with the single judge
One factor for both events in this groupConservative starting value: 45 %In the court’s favour: three times the source value, at most 50 %. The calculator starts with this value.
Standard value from the sources: 15 %There are no official statistics for these errors. As a group they count once, with the highest individual value in the group: 15 out of 100 regional court cases.
Source: Claimant’s tolerance model; events documented by B03, B16, B04
Lawful judgeChange of judge seven days before the hearing
On 29.04.2026, seven days before the hearing, a different judge postponed the hearing by half a year. According to the published court’s allocation plan (Geschäftsverteilungsplan), he is not a member of the competent chamber.
Conservative starting value: 6.0 %In the court’s favour: three times the source value, at most 50 %. The calculator starts with this value.
Standard value from the sources: 2.0 %There are no official statistics for this type of error. The source value is the claimant’s generous assumption in the court’s favour: 2 out of 100 regional court cases.
Source: Claimant’s tolerance model; event documented by B04, B29
Details, legal basis and evidence
On 29.04.2026, seven days before the hearing on 06.05.2026, a judge postponed the hearing “for official reasons” to 04.11.2026. In the published allocation plan (as of 01.02.2026) he is not assigned to the 10th Civil Chamber, but at 2.5/10 to the 6th Civil Chamber and at 7.5/10 to the 2nd Criminal Chamber.
No reason for the change of judge was communicated to us. A later amending decision of the court’s presiding committee (Präsidium) is possible, but we do not have one.
- Date
- 29.04.2026 (hearing was set for 06.05.2026)
- Legal basis
- Lawful judge, Art. 101(1) sentence 2 GG; § 21e GVG
- Evidence
- B04 B29, pp. 89/96/161
- Effect
- to the detriment of the claimant: hearing cancelled, half a year lost
Lawful judgeAllocation of cases remains unexplained
Several enquiries about the basis for the change of judge produced nothing but a reference to the website and to inspection on site.
Conservative starting value: 9.0 %In the court’s favour: three times the source value, at most 50 %. The calculator starts with this value.
Standard value from the sources: 3.0 %There are no official statistics for this type of error. The source value is the claimant’s generous assumption in the court’s favour: 3 out of 100 regional court cases.
Source: Claimant’s tolerance model; event documented by B09, B15
Details, legal basis and evidence
On 04.05., 05.05. and 21.05.2026 we asked about the basis for the change of judge. On 04.06.2026 the administration referred us to the plan published online and, for the plan as of 30.12.2025, to inspection on site:
“… that the current allocation plan is available on the website of the Regional Court of Frankfurt.”Reply of 04.06.2026 (B09) – translation; the German original prevails
Yet it is precisely this plan that does not explain the change. Our letters of 20.10.2026 to the President and to the data protection officer have not yet been answered (deadline 03.11.2026).
- Dates
- Enquiries 04.05., 05.05., 21.05., 20.10.2026 · reply 04.06.2026
- Legal basis
- Verifiability of the lawful judge; §§ 21e, 21g GVG
- Evidence
- B15 B09 B16 B17
- Effect
- to the detriment of the claimant: composition of the court cannot be verified
Case management & timeUrgency without any recognisable weighing
Registration authorities are taking the claimant’s vehicles off the road. The postponements of the hearing show no recognisable weighing against this urgency.
Conservative starting value: 24 %In the court’s favour: three times the source value, at most 50 %. The calculator starts with this value.
Standard value from the sources: 8.0 %There are no official statistics for this type of error. The source value is the claimant’s generous assumption in the court’s favour: 8 out of 100 regional court cases.
Source: Claimant’s tolerance model; event documented by B04, B06, B14
Details, legal basis and evidence
Since February 2026 we have set out the measures by which the registration authorities are taking vehicles off the road. They followed notifications by the insurer under § 51(1) FZV. On 18.05.2026 we also lodged a complaint about delay (Verzögerungsrüge). The postponement decisions of 29.04. and 01.07.2026 show no recognisable weighing against this urgency.
In fairness: on 27.02.2026 the court limited an extension of time for the defendants “in view of the imminent hearing” (B07). On 21.05.2026 it brought the hearing forward on a joint application.
- Dates
- Submission 23.02., complaint about delay 18.05.2026
- Legal basis
- Hearing as early as possible (§ 272(3) ZPO); complaint about delay (§ 198(3) GVG); right of access to justice
- Evidence
- K35.3 K28.8 B14 B04 B06
- Counter-evidence
- B07 B31
- Effect
- to the detriment of the claimant: the damage grows with every month
Group · 4 events · one factor
Time limits, hearing and guidance: months without case management
Four events, one topic: for months there is no setting of time limits, no submissions on the facts by the defendants and no judicial guidance, and the hearing is pushed back by 210 days.
Service & court fileAmended claim: time limit only after 67 days
Our second amendment of the claim of 27.04.2026 went 67 days without the court setting a time limit. On 03.07.2026 the court then said that service from lawyer to lawyer was sufficient.
Details, legal basis and evidence
On 03.07.2026, 67 days after our second amendment of the claim had been transmitted, the court declared that the service from lawyer to lawyer we had effected was sufficient (§ 195(1) sentence 2 ZPO). Only then did it set the defendants a time limit of four weeks. On 21.05.2026 we had expressly asked for a time limit to be set.
In our view, an amendment of the claim with new applications requires prompt case management by the court (§§ 261(2), 270, 273(1) ZPO).
- Dates
- Amendment of the claim 27.04.2026 · time limit set 03.07.2026
- Legal basis
- § 195(1) sentence 2, §§ 261(2), 270, 273(1) ZPO
- Evidence
- B02, pp. 2–3 B13 B21, p. 1
- Effect
- to the detriment of the claimant: more than two months at a standstill
Case management & time135 days to the first statement – without any substance
135 days after our second amendment of the claim, the defendants announced an application to dismiss the claim and reserved “further submissions on the facts”.
Details, legal basis and evidence
The defendants received our second amendment of the claim on 27.04.2026. The court set a time limit only on 03.07.2026; on application it extended it to 09.09.2026. The defendants then announced an application to dismiss the claim and reserved “further submissions on the facts”. The statement does not engage with the study and the expert reports in substance.
The time limit was met. Whether later submissions will still be admitted (§ 296 ZPO) will only be decided at the hearing.
- Period
- 27.04. to 09.09.2026
- Legal basis
- Extension of time §§ 224(2), 225 ZPO; statement of defence §§ 276, 277 ZPO; § 296 ZPO
- Evidence
- B21, p. 1 B02, p. 3 B20 B08
- Effect
- to the detriment of the claimant: time gained without any submissions on the facts
Case management & timeNo substantive guidance after ten months
Around ten months have passed since the claim was filed. To this day there is no substantive guidance from the court on our applications, although we have asked for it.
Details, legal basis and evidence
The claim was received on 30.12.2025. The court has set time limits and hearing dates and on 03.07.2026 gave guidance on procedural questions. Substantive guidance under § 139 ZPO on the merits – that is, on conclusiveness, applications and burden of proof – has not been given to us, around one month before the hearing.
On 27.04. and on 21.05.2026 we expressly asked for such guidance. Judicial guidance must be recorded in the file (§ 139(4) ZPO); its absence can therefore be read from the file.
- Period
- 30.12.2025 to date (hearing 02.12.2026)
- Legal basis
- Substantive case management, § 139 ZPO
- Evidence
- B33, pp. 6, 10–11, 17 B13
- Effect
- to the detriment of the claimant: no opportunity to remedy any shortcomings
Case management & timeHearing pushed back by 210 days
Between the original hearing date (06.05.2026) and the current one (02.12.2026) lie 210 days – across three postponements.
Details, legal basis and evidence
The hearing was moved three times: on 29.04. it was pushed back by 182 days (“for official reasons”), on 21.05. it was brought forward by 70 days on a joint application, and on 01.07.2026 it was pushed back by 98 days because the defendants’ counsel was on holiday. For hearings in July and August, § 227(3) ZPO in principle grants a right to postponement. An exception applies where the proceedings require particular acceleration.
Our reliance on this particular urgency is not discussed in the decisions.
- Dates
- 06.05. → 04.11. → 26.08. → 02.12.2026
- Legal basis
- § 227(1), (3) sentence 3, (5) ZPO; § 216(2) ZPO
- Evidence
- B04 B31 B06 B22, p. 1
- Effect
- to the detriment of the claimant: seven months of delay
One factor for all 4 events in this groupConservative starting value: 30 %In the court’s favour: three times the source value, at most 50 %. The calculator starts with this value.
Standard value from the sources: 10 %There are no official statistics for these errors. As a group they count once, with the highest individual value in the group: 10 out of 100 regional court cases.
Source: Claimant’s tolerance model; events documented by B02, B13, B08, B21, B33, B04, B06, B31
Derivation of the preset standard values & official sources
- No official statistics: Judicial statistics record dispositions and the duration of proceedings, not procedural errors. Every value on this page is therefore a disclosed assumption.
- Source value: the frequency the claimant assumes for each type of error in its litigation dossier (tolerance model) – already generous in the court’s favour, between 2 and 15 out of 100 regional court cases.
- Starting value: three times the source value, at most 50 %. The calculator starts with this. It therefore assumes that each of these errors occurs three times as often as we assume.
- Groups: Related events remain individually visible but count as only one factor in the calculation – with the highest individual value in the group. This way nothing is counted twice.
- Direction of effect: If it were genuinely chance, each error could just as well have gone in the claimant’s favour. The switch accounts for this with ½ per factor; it is initially off.
- The court’s standard (§ 286 ZPO): Judicial conviction requires a degree of certainty that is useful for practical life; circumstantial evidence must be assessed as a whole (BGH, judgment of 07.11.2024 – III ZR 79/23, paras. 32, 39 et seq.).
Orders of magnitude
Where does your result sit in everyday life?
Nobody can picture very large numbers. The ladder places your result among things people know: from rolling a die to winning the Lotto jackpot seven times in a row. Everything in grey is more likely than your result. The ladder follows the sliders.
- 1 : 6Rolling a six
- 1 : 1,024Ten heads in a row when tossing a coin
- 1 : 4,096Twelve heads in a row when tossing a coin
- 1 : 32,76815 heads in a row when tossing a coin
- 1 : 262,14418 heads in a row when tossing a coin
- 1 : 529,221Your result
- 1 : 649,740Royal flush when dealt five cards4 of 2,598,960 possible hands
- 1 : 14 millionAll six numbers right in the German Lotto 6/49
- 1 : 140 millionGerman Lotto jackpot: six numbers plus the bonus number (Superzahl)
- 1 : 8.2 billionBlindly picking one particular person from everyone on EarthWorld population around 8.2 billion (UN World Population Prospects 2024)
- 1 : 1.1 trillion40 heads in a row when tossing a coin
- 1 : 19.6 quadrillionWinning the German Lotto jackpot twice in a row
- 1 : 435 quadrillionBlindly guessing the one right second since the Big Bang13.787 billion years ≈ 4.35 · 10¹⁷ seconds (Planck 2018)
- 1 : 1.18 sextillion70 heads in a row when tossing a coin
- 1 : 2.73 septillionWinning the German Lotto jackpot three times in a row
- 1 : 7 octillionBlindly hitting one particular atom in your own bodyEstimate: around 7 · 10²⁷ atoms in an adult
- 1 : 382 nonillionWinning the German Lotto jackpot four times in a row
- 1 : 1.33 undecillion120 heads in a row when tossing a coin
- 1 : 53.5 duodecillionWinning the German Lotto jackpot five times in a row
- 1 : 1.43 quattuordecillion150 heads in a row when tossing a coin
- 1 : 7.48 quindecillionWinning the German Lotto jackpot six times in a row
- 1 : 133 quindecillionBlindly hitting one particular atom of the entire EarthEstimate: around 1.3 · 10⁵⁰ atoms
- 1 : 1.05 octodecillionWinning the German Lotto jackpot seven times in a row
Method & sources
How it is calculated – and where the model has its limits
Standard value per point
There are no statistics on procedural errors: judicial statistics count cases and their duration, not errors. In their place we use deliberately generous values in the court’s favour.
Multiply
If the seven factors were coincidences independent of one another, the probability that all of them occur in the same case would be the product of the individual values.
Direction (optional)
An undirected error would hit sometimes one side, sometimes the other. Anyone who puts this at ½ per factor multiplies additionally by (½)⁷ = 1/128. The switch is initially off.
Conservative starting values (7 sliders): 1.89 · 10⁻⁶ → 1 : 529,221
Source values: 8.64 · 10⁻¹⁰ → 1 : 1.16 billion
All sliders at 0.1 %: 1 · 10⁻²¹ → 1 : 1 sextillion
| No. | Error complained of | Legal basis | Source value | Conservative | Basis of the source value |
|---|---|---|---|---|---|
| 01 | Tracking histories: not requested, not in the court file One factor for both events in this group | Completeness of the file; §§ 182(3), 298a, 299(1) ZPO; Art. 103(1) GG; § 286 ZPO; § 418(2) ZPO | 4.0 % | 12 % | Claimant’s tolerance model; events documented by B01, B02 |
| 02 | Second refusal – with the charge of a “shot in the dark” | Art. 103(1) GG; prohibition of anticipatory assessment of evidence | 3.0 % | 9.0 % | Claimant’s tolerance model; event documented by B02 |
| 03 | Composition: single judge instead of chamber One factor for both events in this group | § 348(3) sentence 1 no. 2, § 348a(1) no. 2 ZPO; Art. 101(1) sentence 2 GG; § 348(3) sentence 1 no. 2 ZPO (grounds for referral); Art. 101(1) sentence 2 GG | 15 % | 45 % | Claimant’s tolerance model; events documented by B03, B16, B04 |
| 04 | Change of judge seven days before the hearing | Art. 101(1) sentence 2 GG; § 21e GVG | 2.0 % | 6.0 % | Claimant’s tolerance model; event documented by B04, B29 |
| 05 | Allocation of cases remains unexplained | Art. 101(1) sentence 2 GG; §§ 21e, 21g GVG | 3.0 % | 9.0 % | Claimant’s tolerance model; event documented by B09, B15 |
| 06 | Urgency without any recognisable weighing | § 272(3) ZPO; § 198 GVG; Art. 2(1) in conjunction with Art. 20(3) GG; Art. 6(1) ECHR | 8.0 % | 24 % | Claimant’s tolerance model; event documented by B04, B06, B14 |
| 07 | Time limits, hearing and guidance: months without case management One factor for all 4 events in this group | §§ 195(1) sentence 2, 261(2), 270, 273(1) ZPO; §§ 224(2), 225, 276, 277 ZPO; equality of arms (Art. 3(1) in conjunction with Art. 20(3) GG); § 139(1), (4) ZPO; § 273(1) ZPO; § 216(2), § 227(1), (3), (5) ZPO | 10 % | 30 % | Claimant’s tolerance model; events documented by B02, B13, B08, B21, B33, B04, B06, B31 |
The honest objections
A calculation is only as good as its assumptions. These objections are justified, and this is how the page deals with them:
“The errors are not independent.”
Probably not, and that is exactly the point. That is why three groups of related events count as only one factor each. Nos. 6 and 7 are also closely linked: set the slider for no. 7 to 100 % and you take that one out of the calculation as well. If chance practically rules out the coincidence, a common cause becomes more plausible. Which one, the calculation does not say: overload and staff changes are just as possible.
“The points were selected after the fact.”
True. Anyone searching with hindsight will always find something rare. That is why every value is disclosed, the calculator starts with three times the source values, and every slider goes up to 100 %.
“The 50:50 direction does not hold.”
A fair point: whoever complains mainly complains about what burdens them, and delays naturally hit the side bringing the claim. That is why the direction factor is a switch and off by default.
“Is each of these really an error?”
That is an open legal question. These are the claimant’s complaints; the court and the defendants see some things differently. For nos. 4 and 5 we have requested the documents on the allocation. Where the file also speaks in the court’s favour, this is stated under the respective number.
What the number means – and what it does not
The number states how rare the coincidence would be under your assumptions if only independent chance were at work. It is not the probability that someone acted intentionally. Anyone who infers intent directly from a very small probability of chance commits a well-known fallacy. The calculation says something else: the explanation “nothing but mishaps independent of one another” fits the court file poorly. Which explanation fits better is for the proceedings to establish.
The estimates are sensitive: if every single estimate is off by a factor of 3, with seven factors the result shifts by more than a factor of two thousand (3⁷ = 2,187). That is why the calculator shows the order of magnitude, not an exact number. And that is why the calculator starts with three times the source values (at most 50 %). Even then the coincidence is rarer than 18 heads in a row.
Sources
- Age of the universe: 13.787 ± 0.020 billion years, Planck Collaboration, Planck 2018 results VI, Table 2.
- German Lotto 6/49: 1 : 13,983,816 for all six numbers right, 1 : 139,838,160 with the bonus number (Superzahl) (calculated exactly: 49 choose 6, times 10). Eurojackpot happens to have the same odds (50 choose 5 times 12 choose 2).
- Royal flush: 4 of 2,598,960 possible five-card hands = 1 : 649,740. Coin toss: n heads in a row = 1 : 2n.
- World population around 8.2 billion: UN World Population Prospects 2024. Atoms in the human body (around 7 · 10²⁷) and in the Earth (around 1.3 · 10⁵⁰): orders of magnitude derived from mass and mean atomic mass.
- Statutes (German): ZPO, GVG, GG, FZV.
- Evidence references (B…, K…, BLD…) refer to our exhibits and letters in the proceedings; details are on Rule of law and in the Timeline. Content from our inspection of the court file is rendered in our own words.