On average, half of all Bitcoin bought at cycle lows was sold within just 210 days. For 2022 bottom buyers, that meant selling around $26k, before the real rally had even started. Every cycle back to 2011 shows the same pattern, and it says a lot about how today's bottom buyers are likely to behave👇🏼

presets are now also available in the sidebar, if you were a user of bitview.space you should feel right at home studio.bitview.space

presets now browse-able in "All charts" studio.bitview.space

from a bad prototype to this, in 7 days ! we truly live in crazy times where the only limit is your imagination (and your tokens)

ok folks drawings and folders ARE IN studio.bitview.space

you can now import any chart from bitview.space in the studio 🎉 next up: drawing studio.bitview.space

first version has finished cooking introducing... bitview studio: preview edition studio.bitview.space

cooking like a freakin madman 👨‍🍳 look at those mini chart logos

For the longest time my definition of a Long Term Holder was to simply follow the industry standard: 155 days, which was set by @glassnode in 2020 (iirc) Although, I did round it down to 5 months/150 days, to make my life, as a builder, much easier While I was certain there had to be more to it, I couldn't find the time to study it properly and so I was delighted was I first saw @MattCrosbyPro put out a very thorough study where he tried to recreate @glassnode 's findings and draw his own conclusions from his own analysis tldr: the threshold changes constantly (110 days to 280 days iirc) and he settled on a dynamic formula that changes over time I personally think that a dynamic version is maybe a bit too complicated, jumpy and potentially laggy (in the sense that it will react after the fact) and so instead my thinking was that since few things in life are actually binary, a better framework is to think of it as a transition period: 4 to 6months (and arguably 3 to 7 months), where the hodler may be both How do you illustratre that is a different question, I've been using min-max clouds (decent to draw a buffer zone and display the uncertainty) of the ranges: <4/<5/<6m for sth and >4/>5/>6m for lth but there are many other ways But a threshold or a range is not enough, when you think about it: a holder holds coins for a reason. They will eventually do something with it, wether that's spending, lending, etc... otherwise those are just dead coins and they don't matter, worse they pollute the dataset when we think of actual hodlers So now the question becomes: how do we quantify which part of the supply matters and which doesn't ? That's how I ended up with two weighting systems that I initially created for the bedrock model: - wakefulness (extended cointime, with the cointime framework created by @_checkonchain and @dpuellARK): cumulative coin-days consumed ÷ cumulative coin-days created - mobility (coinflow): estimated chance a coin will eventually be spent, based on historical spending rates both of which are computed independently for each age range cohort (up to 1h, 1h to 1d, 1d to 1w, 1w to 1m, etc..) which also means the finer you go, the more truthful the result will be The question that both try to answer is: what part of the supply is relevant and the supply that's most likely to ever move ? Both roughly agree and neither is perfect, it's just probabilities but that's the best system I got at the moment To give an example today: - mobility gives a 100% chance that the younger than 1h will ever move, 89.6% for the 5 to 6 months, 13.2% for 8 to 10 years, and 0.4% to the older than 15y supply - wakefulness gives respectively 99.5%, 82.1%, 11.4% and 0.1% Which means that when we model things out we can ignore more than 99.6% of the supply older than 15y, that's how much pollution the unweighted lth metrics carry with them Lastly, the metrics that you look at, are even more important than the cohorts themselves. No one ever thinks in coin terms and I would argue that as soon as the coins had a price tag, no one ever did. Instead folks think in dollar terms, how much they have, how much they're up (greed) and how much they're underwater (pain) Which means that metrics like the realized price (the average cost basis per coin, or the average price paid for each coin) isn't useful A better metric, which I made a while ago, is the capitalized price (the average cost basis per coin, or the average bitcoin price at which dollar was invested at) which is a much better representation of the average mood, sentiment or emotion Concrete example: take 2 coins, 1st bought at 10k, 2nd at 100k realized price: (10k + 100k) / 2 = $55k capitalized price: (10k^2 + 100k^2) / (10k + 100k) ~= $91.8k While the realized price shows the equilibrium at which both are as much as much up as down, the capitalized shows where the money was actually invested, where is the point where the average dollar invested is either in profit or in loss, the person that invested at 100k is 10 times more invested than the person at 10k Bringing this back to long term holders, we never touched the long term holders realized price, nor the normal realized price for that matter (and we never will again, unless maybe quantum), does that mean that the average long term holder was happy ? No absolutely not, so much money was invested at the top range of the bull, at $58k, 93% of the capital invested was in loss, or 83% of the lth capital invested, so clearly there was pain and the capitalized price shows that, we're just above the unweighted lth cap price at $84k, and just below the wakeful lth cap price at $88.7k If you sum it all up, you can get a pretty decent story, and that's what on-chain data is all about, an incredible lens at what folks are feeling and doing

hello world