# Last edited on 2014-04-29 23:06:16 by stolfilocal [b]Chinese Slumber Method prediction for Wednesday April 30[/b] [b]Prediction valid for:[/b] Wednesday 2014-04-30, 19:00--19:59 UTC (not before, not after) [b]Huobi's predicted price:[/b] 2720 CNY [b]Bitstamp's predicted price:[/b] 439 USD [url=http://www.ic.unicamp.br/~stolfi/temp/2014-04-29-prediction-check-HUBI-annotated.png][img width=600]http://www.ic.unicamp.br/~stolfi/temp/2014-04-29-prediction-check-HUBI-annotated.png[/img][/url] [url=http://www.ic.unicamp.br/~stolfi/temp/2014-04-29-prediction-check-BSTP-annotated.png][img width=600]http://www.ic.unicamp.br/~stolfi/temp/2014-04-29-prediction-check-BSTP-annotated.png[/img][/url] [ [url=http://www.ic.unicamp.br/~stolfi/temp/2014-04-29-prediction-check-legend.txt]Plot legend[/url] ] Today's data point, like yesterday's, was fairly good (S = 0.0035, W = 0.778), and not very far from the previously assumed trend line. Again two choices for the trend seemed reasonable, the straight line determined by the last two points only, or a straight trend fitted by weighted least squares to the last five points (which would ignore yesterday's given its low weight). I chose the latter, but multiplied the Slumber Weight W by a "remoteness" weight WF = K**(DF-d) where K = 0.8 and (DF-d) is the number of days before the last data point (Apr/29). The fitted line was A + B*(d-d0), where d-d0 is the number of days since Apr/25, A = 2813.92 and B = -18.83. The Bitstamp prediction, as usual, is the Huobi prediction divided by the currency conversion factor R, which was assumed to be 6.20 CNY/USD. It was 6.25 at today's Slumber time, and 6.05, 6.09, and 6.05 for the Slumber Times at Apr/28, Apr/26, and Apr/25. [b]Checking the previous prediction[/b] Prediction was posted on: Tuesday 2014-04-29, 02:44 UTC Prediction was valid for: Tuesday 2014-04-29, 19:00--19:59 UTC (~16 hours later) The prediction for Huobi was rather off: Huobi's predicted price: 2684 CNY Huobi's actual price (L+H)/2: 2761 CNY Error: 77 CNY (~12 USD) The prediction for Bitstamp was almost exact, but only because the R factor changed: Bitstamp's predicted price: 443 USD Bitstamp's actual price (L+H)/2: 442 USD [b]Error: 1 USD[/b] [b]NOTE[/b]: "Man differs from the animal only by a little; most men throw that little away." -- Confucius