HellenicSlavotolian (@HellenicSlav) header
HellenicSlavotolian (@HellenicSlav) profile picture

Hellenic-Slavotolian

Peloponnisos, Greece
Joined October 2026
Photo by HellenicSlavotolian (@HellenicSlav)Photo by HellenicSlavotolian (@HellenicSlav)Photo by HellenicSlavotolian (@HellenicSlav)Photo by HellenicSlavotolian (@HellenicSlav)Photo by HellenicSlavotolian (@HellenicSlav)Photo by HellenicSlavotolian (@HellenicSlav)Photo by HellenicSlavotolian (@HellenicSlav)Photo by HellenicSlavotolian (@HellenicSlav)Photo by HellenicSlavotolian (@HellenicSlav)Photo by HellenicSlavotolian (@HellenicSlav)
HellenicSlavotolian retweeted
qpAdm models don't mean much when it is an inconsistent tool that can yield wildly different results if you play around with the reference populations. Keeping the same source populations but changing the reference populations can turn a passing model into a failing one, or a model with incredibly different proportions of the same source populations. G25 is way more consistent since all samples, modern or ancient, are based off the same 25 PCs aka the "reference populations". It avoids this layer of inconsistency and variability that qpAdm cannot avoid, by it's own design. Also the screenshots are relevant, the models and distance differences are replicable, calling the tool a hobbyist tool and appealing to authority (once again) does not make it worse and finally if you want to appeal to authority I want to remind you that G25 was created by David Wesolowski (Davidski) who is an independent researcher that has taken part in different peer reviewed studies such as the recent Albanian paper. We know the ancestry was not from "Southeast Anatolia" because Southeast Anatolia is, again, not a real historical region but something that Greek academia made up to cope with the clear West-Central Anatolian-shift and Near Eastern influx in Europe during the Roman-era supported by aDNA, admixture modeling, Y-DNA lineages and also historical evidence. Southeast Turkey is not Anatolia and never was. Do I really need to teach you the geography and history of a region which your people have claimed since forever? They're obviously Anatolians based on their geographic location, time period, Neolithic admixture, PCA plot and (when possible) IBD matches and YDNA.
2
7
1
26
1,122
HellenicSlavotolian retweeted
Laza-Liar-idi 😆
🤖 Made with AI
Calling this AI-generated while you yourself openly use Claude for map-checking and half your feed is a nonstop carousel of AI-assisted nationalist continuity copes is pure projection. Your 2017 paper is a methodological embarrassment that nationalists still wave around as “proof.” It showed nothing more than Class I similarity: PCA position + a handful of f4-statistics indicating that modern Greeks “resemble” Mycenaeans after some Neolithic dilution. No qpAdm with modern Greek targets, no formal 2-/3-source models, no IBD/segment-sharing tests, n=19 ancient individuals. “Continuity but not isolation” is an unfalsifiable slogan, not a result. Detailed breakdown here: teapawt.net/AlbertusMvagnus/status… Demanding a 3-source model while your preferred Imperial Roman source already carries Aegean-related ancestry (something you yourself say “must be demonstrated”) is circular special pleading. A fitting 2-source model without Mycenaean is exactly the result one expects if the Mycenaean component is small or poorly differentiated. “Continuity with change” has become the rhetorical shield for every failed model that fails to recover the ancestry you want. Show the same configuration across multiple targets with proper outgroups and SEs, or stop treating 2017 as holy writ.
4
5
31
1,081
HellenicSlavotolian retweeted
Fraudstertzaki:
Calling this AI-generated while you yourself openly use Claude for map-checking and half your feed is a nonstop carousel of AI-assisted nationalist continuity copes is pure projection. Your 2017 paper is a methodological embarrassment that nationalists still wave around as “proof.” It showed nothing more than Class I similarity: PCA position + a handful of f4-statistics indicating that modern Greeks “resemble” Mycenaeans after some Neolithic dilution. No qpAdm with modern Greek targets, no formal 2-/3-source models, no IBD/segment-sharing tests, n=19 ancient individuals. “Continuity but not isolation” is an unfalsifiable slogan, not a result. Detailed breakdown here: teapawt.net/AlbertusMvagnus/status… Demanding a 3-source model while your preferred Imperial Roman source already carries Aegean-related ancestry (something you yourself say “must be demonstrated”) is circular special pleading. A fitting 2-source model without Mycenaean is exactly the result one expects if the Mycenaean component is small or poorly differentiated. “Continuity with change” has become the rhetorical shield for every failed model that fails to recover the ancestry you want. Show the same configuration across multiple targets with proper outgroups and SEs, or stop treating 2017 as holy writ.
2
4
20
594
Fraudstertzaki:
Calling this AI-generated while you yourself openly use Claude for map-checking and half your feed is a nonstop carousel of AI-assisted nationalist continuity copes is pure projection. Your 2017 paper is a methodological embarrassment that nationalists still wave around as “proof.” It showed nothing more than Class I similarity: PCA position + a handful of f4-statistics indicating that modern Greeks “resemble” Mycenaeans after some Neolithic dilution. No qpAdm with modern Greek targets, no formal 2-/3-source models, no IBD/segment-sharing tests, n=19 ancient individuals. “Continuity but not isolation” is an unfalsifiable slogan, not a result. Detailed breakdown here: teapawt.net/AlbertusMvagnus/status… Demanding a 3-source model while your preferred Imperial Roman source already carries Aegean-related ancestry (something you yourself say “must be demonstrated”) is circular special pleading. A fitting 2-source model without Mycenaean is exactly the result one expects if the Mycenaean component is small or poorly differentiated. “Continuity with change” has become the rhetorical shield for every failed model that fails to recover the ancestry you want. Show the same configuration across multiple targets with proper outgroups and SEs, or stop treating 2017 as holy writ.
2
4
20
594
HellenicSlavotolian retweeted
Calling this AI-generated while you yourself openly use Claude for map-checking and half your feed is a nonstop carousel of AI-assisted nationalist continuity copes is pure projection. Your 2017 paper is a methodological embarrassment that nationalists still wave around as “proof.” It showed nothing more than Class I similarity: PCA position + a handful of f4-statistics indicating that modern Greeks “resemble” Mycenaeans after some Neolithic dilution. No qpAdm with modern Greek targets, no formal 2-/3-source models, no IBD/segment-sharing tests, n=19 ancient individuals. “Continuity but not isolation” is an unfalsifiable slogan, not a result. Detailed breakdown here: teapawt.net/AlbertusMvagnus/status… Demanding a 3-source model while your preferred Imperial Roman source already carries Aegean-related ancestry (something you yourself say “must be demonstrated”) is circular special pleading. A fitting 2-source model without Mycenaean is exactly the result one expects if the Mycenaean component is small or poorly differentiated. “Continuity with change” has become the rhetorical shield for every failed model that fails to recover the ancestry you want. Show the same configuration across multiple targets with proper outgroups and SEs, or stop treating 2017 as holy writ.
This is a good example of the kind of AI-generated analysis that I warned about in an earlier post. The poster is trying to prove the lack of Mycenaean ancestry, yet his model does not have Mycenaean on the left. To disprove such ancestry one would test the 3-source model and...
4
8
3
42
5,529
HellenicSlavotolian retweeted
Hahahahahahah ahahahahah hahahah cant stop Laughing 😭😭😭😭😭😭👇👇👇
Southeast Europe, Anatolia and the Levant on a regional PCA (PC1 × PC2). Athens and Thessaloniki Greeks, modern Turks, Bulgarians, Byzantine Anatolia samples and Levantine groups (Lebanese Christian/Muslim, Druze, Palestinian). The plot shows the intermediate position of the northern Greek samples between Balkan and Anatolian clusters.
1
1
1
12
526
Hahahahahahah ahahahahah hahahah cant stop Laughing 😭😭😭😭😭😭👇👇👇
Southeast Europe, Anatolia and the Levant on a regional PCA (PC1 × PC2). Athens and Thessaloniki Greeks, modern Turks, Bulgarians, Byzantine Anatolia samples and Levantine groups (Lebanese Christian/Muslim, Druze, Palestinian). The plot shows the intermediate position of the northern Greek samples between Balkan and Anatolian clusters.
1
1
1
12
526
HellenicSlavotolian retweeted
Pete. 🤣🤣🤣🤣🤣🤣🤣The real descendant of Leonida, ahu, ahu, ahahahahhahahahha.
2
1
3
51
HellenicSlavotolian retweeted
Replying to @Theodoros2A5
Theodoros: Two days of “garbage”, “slop”, “crackpottery”, “scam”. I counted: zero models named, zero numbers, zero errors shown. Impressive for someone who presents himself as an expert. To make it easier for you, here is the model you find so terrible: Poland_EarlyMedieval_Slav + Turkey_Byz_SW2 on Greek_Thessaloniki, 36.6% / 63.4%, SE 0.028, p = 0.11, run by hand in RStudio with the same code you use. The Mycenaeans add 3.7% ± 9.8%, so nothing. Show me the error. In case getting started is hard, here are a few handholds that any specialist could answer without effort: Which proxy is wrong, and why? Which right population is missing or disturbing the fit? Is Byz_SW2 too small with 2 individuals? (Yes, I say so myself. You could have spotted it.) Which number would you expect instead? I’m waiting, and so is everyone reading along. “You don’t understand the tool” sounds like an argument as long as you don’t have to say anything concrete. You only have to name one number. Until then, two days of insults leave exactly one thing: the tone.
3
1
1
7
496
HellenicSlavotolian retweeted
Whether I enter a model by hand in RStudio or let a tool run several in a row makes no difference to the result. qpAdm runs the same function (admixtools::qpadm) on the same dataset (AADR v66.p1, 1240K) either way. That is an objection to the tool, not to any result. So I re-ran the models by hand in RStudio, with identical results. Example: Poland_EarlyMedieval_Slav + Turkey_Byz_SW2 gives 36.6% and 63.4% for Greek_Thessaloniki (SE 0.028 each), p = 0.11, on 405,640 shared SNPs. Adding Mycenaean_HQ7 keeps it at p = 0.056, and the Mycenaeans come out at 3.7% ± 9.8%, indistinguishable from zero. Drop either of the two main sources and the model collapses (p = 3e-127 without the Slavs, p = 1.5e-21 without the Byzantines). Every single model, every right-population list and every run can be checked, and I can show you the code for each one. So the finding stands: the Thessaloniki Greeks are statistically compatible with a mixture of Slavic and Anatolian-Byzantine ancestry. If you want to attack the model, attack the proxies, the rights or the sample. That is scientific. “The AI did it” is not. And one more thing: your nationalism changes nothing about the data. The results stand whether you like them or not, and anyone can recompute them. Anyone who has no substantive objection to the models and hangs on how the tool is operated only embarrasses himself.
Replying to @Theodoros2A5
you might have some idea about nationalism with racist thoughts, but you have nothing to do with genuine research, you belong deported from our beautiful Germany
2
6
2
14
3,056
HellenicSlavotolian retweeted
Slavotolic Greeks would do and say anything, absolutely anything, except ever tell the truth.
🤖 Made with AI
Whether I enter a model by hand in RStudio or let a tool run several in a row makes no difference to the result. qpAdm runs the same function (admixtools::qpadm) on the same dataset (AADR v66.p1, 1240K) either way. That is an objection to the tool, not to any result. So I re-ran the models by hand in RStudio, with identical results. Example: Poland_EarlyMedieval_Slav + Turkey_Byz_SW2 gives 36.6% and 63.4% for Greek_Thessaloniki (SE 0.028 each), p = 0.11, on 405,640 shared SNPs. Adding Mycenaean_HQ7 keeps it at p = 0.056, and the Mycenaeans come out at 3.7% ± 9.8%, indistinguishable from zero. Drop either of the two main sources and the model collapses (p = 3e-127 without the Slavs, p = 1.5e-21 without the Byzantines). Every single model, every right-population list and every run can be checked, and I can show you the code for each one. So the finding stands: the Thessaloniki Greeks are statistically compatible with a mixture of Slavic and Anatolian-Byzantine ancestry. If you want to attack the model, attack the proxies, the rights or the sample. That is scientific. “The AI did it” is not. And one more thing: your nationalism changes nothing about the data. The results stand whether you like them or not, and anyone can recompute them. Anyone who has no substantive objection to the models and hangs on how the tool is operated only embarrasses himself.
4
5
34
1,230
HellenicSlavotolian retweeted
When pseudo-Greeks try to measure themselves against Albanians, cock-sucking is a foregone conclusion.
🤖 Made with AI
How Greeks feel when they meet an Albanian after getting their DNA test results (80% 🇬🇷 Asian).
1
3
1
14
741
When pseudo-Greeks try to measure themselves against Albanians, cock-sucking is a foregone conclusion.
🤖 Made with AI
How Greeks feel when they meet an Albanian after getting their DNA test results (80% 🇬🇷 Asian).
1
3
1
14
741
HellenicSlavotolian retweeted
Replying to @KingottoofGreec
Brutal!! I too found this idiot both an insufferable bore and completely devoid of intelligence. Kudos @nntaleb
2
4
77
HellenicSlavotolian retweeted
You sre also pro albanian, you hate greece, you preof greeks in Albania are Gypsis 😅👇👇
1
4
55
HellenicSlavotolian retweeted
Are you retarded? I've made it abundantly clear my pride in my majority Alashiyan ancestry, only to point out the irony, Cypriots have as much (more than most in fact, per Olalde et al 2023) Mycenaean than mainlanders/islanders alike. This due to direct migration post Helladic. To point out truth, that Dorians were more steppe heavy like Logkas samples and originated in Epirus is to state dispassionate fact. The bulk of modern mainland genes are Slavic and invariant of outgroups and references. The population exchange of Anatolia skewed in these Slav heavy 'Thessalonians' a more BA Anatolian affinity which without adequate sources in qpadm are wrongly inferred (or intentionally so) as BA Mycenaean. I don't subscribe to the naïve postmodern orientalism trope, only that it was the Semitic Civilisations that were producing works of the sciences, arts far preceding those of Europe. teapawt.net/ChristosA89/status/211…
Yes, agreed. The premise of qpAdm is F4 statistics to differentiate via linear algebra shared drift, not only calculate shared allele frequencies, as does PCA or fst. My point, even these BA Mycenaean and BA (west) Anatolian samples can, with the correct outgroups, be distilled. In the case of Mycenaeans it is the extra Catacomb component. For instance, I as a Greek Cypriot, score ~30% Logkas Mycenaean and the rest Bronze Age Canaanite. It is imperative to use the correct sources, as I did per Cypriot historicity, or risk inflating values. Northern Greeks equally score around the ~30% mark Mycenaean, so to then inaccurately inflate their proportions due to Roman Anatolian is not only statistically dishonest but illiterate. What Cypriots contain as native EBA Alashiyan, the mainlanders gained in Slavic of identical coefficients.
3
3
1
4
248
HellenicSlavotolian retweeted
They’re done, more copium for our Slavotolians pls 😆
🤖 Made with AI
1/ [rescaled maps] On the entropy-rescaled PCA (Taleb et al.), an independent method, Thessaloniki’s nearest mean is Bulgarian and Athens’s is Turkish: a Slavic-related pull on Thessaloniki in both the ordinary and the rescaled maps. Byzantine West Anatolia sits nearer Athens and Crete there, so I claim no more than that. Positions, not ancestry shares. 2/ [Fig. 5] On a PCA where 41 Byzantine Anatolians help build the axes, they sit with Turks and Lebanese, Athens with Turks, Thessaloniki toward Bulgarians. Mainland Greeks also don’t sit at the Mycenaean position (Mycenaean PC2 ≈ 0.041, Athens ≈ 0.007, Thessaloniki ≈ 0.002). 3/ “Same individuals = the ones Olalde et al. used” is a pointer to a paper, not a list. Which individuals in which source, which outgroups, which SNP set, which p-values? Post them as text or a file. I re-ran M1–M4 from your descriptions, all public: M1 infeasible, M2 and M3 rejected, M4 (Kaman) passes but fails my pre-declared checks. 4/ Without Kaman: Slavic + Imperial Roman Turkey (n=4) fits Thessaloniki on R10 at p=0.52 (Kaman: 0.071). Other Roman-era Anatolian groups also fit on R10, and Slavic + Samantaş (n=2, Byzantine) is the only non-rejected model among 24 with Byzantine sources. Caveats: R10 only (rejected on RB13), small groups, and a fit is the absence of rejection, not proof of descent. Slavic, Anatolian, Caucasian and Levantine contributions are a plausible reading; the maps can’t separate them. Post your lists and I rerun them in public.
17
12
1
73
4,771
HellenicSlavotolian retweeted
Replying to @Theodoros2A5
I ran your configuration on the Olalde panel (his sources, 13 outgroups): all of your Anatolian variants (Alalakh, Arslantepe, Old Hittite/Kaman) are rejected, closest is Kaman/Thessaloniki at p = 0.041. More importantly, I re-checked ~250 feasible models from my earlier runs. The pattern is the same in all of them: Slavic is the most stable component. Median SE 2.3%, weight 18–42% in every accepted model. All 22 accepted models (p ≥ 0.05, SE < 10%) contain a Slavic source. Anatolian is second. Its weight depends on which Anatolian source you pick, but it stays well determined (median SE ~3%). Mycenaean is the forced component. Its weight swings from 0.3% to 94% depending on the other sources (19–82% in accepted models), median SE 6%. Only 7 of 22 accepted models include it at all. Slavic + Mycenaean without an Anatolian source is rejected in every case (p < 1e-16 to 3e-11). A fit is not descent, so none of this proves a specific Mycenaean share. But it shows the Mycenaean share is not identified by these data. Models and links on request.
1
1
1
8
1,036
HellenicSlavotolian retweeted
The Slavotoolian Geneticist Army 😅
Theodoros, “crook”, “liar”, “scammer”: three insults, zero numbers. That’s the third reply in a row without a model. I re-ran your M1–M4 from your own descriptions, all public. M1 infeasible (p=4e-5), M2/Alalakh rejected (p=1.8e-5), M3/Arslantepe rejected (p=2.2e-11). Only M4/Kaman passes (p=0.071), and my published pre-declared checks mark that three-source fit invalid (fragile, fails for Athens). Your “100,000 SNPs” rule has no citation. I re-ran the Slavic + Anatolian models with Byzantine sources at 400k+ SNPs. You didn’t address that. You say you shared individuals and outgroups. Link the exact post. I can only find “same individuals, same outgroups” as a sentence. Give me the IDs and I’ll rerun your exact setup and publish it. “AI did it” isn’t a rebuttal. Fit rule p ≥ 0.05, weights in [0,1], SE < 0.10, every run open to inspection. Post a Thessaloniki model that fits at least as well as the Samantaş one, with IDs, outgroups and SNP set. Otherwise it’s just insults. nexju.com/share/6X02menlpMpR…
2
4
35
1,041
The Slavotoolian Geneticist Army 😅
Theodoros, “crook”, “liar”, “scammer”: three insults, zero numbers. That’s the third reply in a row without a model. I re-ran your M1–M4 from your own descriptions, all public. M1 infeasible (p=4e-5), M2/Alalakh rejected (p=1.8e-5), M3/Arslantepe rejected (p=2.2e-11). Only M4/Kaman passes (p=0.071), and my published pre-declared checks mark that three-source fit invalid (fragile, fails for Athens). Your “100,000 SNPs” rule has no citation. I re-ran the Slavic + Anatolian models with Byzantine sources at 400k+ SNPs. You didn’t address that. You say you shared individuals and outgroups. Link the exact post. I can only find “same individuals, same outgroups” as a sentence. Give me the IDs and I’ll rerun your exact setup and publish it. “AI did it” isn’t a rebuttal. Fit rule p ≥ 0.05, weights in [0,1], SE < 0.10, every run open to inspection. Post a Thessaloniki model that fits at least as well as the Samantaş one, with IDs, outgroups and SNP set. Otherwise it’s just insults. nexju.com/share/6X02menlpMpR…
2
4
35
1,041